I went as far as to order a frame from Tsunami Bikes that I thought would fulfill some of the things I thought important. I couldn't get integrated brakes on my budget, nor could I get some cool faired stem-fork deal, but I could do a few things.
First I got narrow tubing. I later realized that I have such a small frame that the tubing really wouldn't make that much difference but if I was going to get an aero frame I wanted the tubing to be narrow.
Second I got a rear wheel cutout. It may not make a huge difference with a direct headwind - in fact I highly doubt it - but it should make for a better sail in a crosswind. I figured that in a cross tailwind it'd be even better. As a bonus I could get really short chainstays because I wouldn't be limited to what room the tube left the rear tire.
Thirdly I got internal cable routing. I dealt with such things enough in the bike shop days so routing cables and such isn't such a pain (although I probably wouldn't have agreed with you the first time I installed a front derailleur cable. The shifter cables travel through the downtube and the rear brake through the top tube. I had no "through fork" method for doing the front brake cable (like for BMX bikes) so that cable hangs out in the breeze.
(I have to admit the main reason I got internal cables is that it's much, much, much easier to clean. I read somewhere that a foot of exposed cable housing costs about one watt. With my very small frame I have maybe 2 watts of drag due to cable housing. That's not a lot. If I had a 35 inch inseam I might have approached things a bit differently.)
Fourth I got an integrated seat post (or ISP). I knew and forgot that BH bikes "invented" it. I think it has its place but what I have is more of a style thing. I think that a shaped ISP has its place, able to significantly reduce shock while presenting a minimal profile to the wind. My ISP does neither but it looks cool so I did it.
Finally I had the frame built with only secondary consideration for the water bottles. I knew that my frame thoughts gave me only "fine" benefits. This is significant so I'll explain why.
In the early aero days things were so conservative that a 16 mm tall rim was considered "aero". Seriously! Nowadays a 32 mm rim would be called "really conservative". When disk wheels and then tall profile rims burst onto the scene they gave tremendous benefits over things like a 16 mm tall "aero" wheel. You could save many minutes in a 40 km TT by using a true aero wheel like a Specialized TriSpoke (with HED selling the wheel, with a slightly modified rear hub, as the HED3)
I call that a "coarse benefit". What's the point of quibbling over a second or two when you can save 4 or 5 minutes? Suddenly aero brake cables (some were still being routed up in the air) seemed kind of minor. Aero wheels would make a huge difference.
I realized this a long time ago and used the TriSpokes in races where I expected the pace to be flat out all the time. In tight crits I would use my normal superlight 28 spoke 280/330g tubulars, but by the mid 90s I had a slew of aero wheels at my disposal - the TriSpokes, Zipp 440s (pre-404s), Zipp 340s (pre-340s), Spinergy Rev-Xs, a rear disk wheel (17 mm wide rim prototype and extremely light), Campy Ventos (all aluminum rim aero wheels - they were tanks), and finally my first aero wheelset, the 16 mm tall Araya ADX-4s.
Not everyone seemed so openminded, especially at the pro level. That held true for a long time. Even in the mid-90s most pros didn't use significantly aero wheels. Spinergy made some inroads with the Mapei, Rabobank, and Polti teams, with some very, very good riders on the wheels. Next to those wheels the relatively primitive, full aluminum Campy wheels looked heavy and outdated.
Then Mavic hit the pro ranks with the Cosmic. Aero wheels were still not totally accepted but some of the better Classics racers used the Cosmics to great effect. Well that and doping apparently, but still, if they were all doping then the aero wheels would help make a difference.
Finally Zipp hit the big time with CSC. After a year or two of disasters (wheels breaking on the cobbles usually) Zipp could boast some major tough victories on their wheels.
Now everyone uses aero wheels. Not only that, everyone uses pretty tall aero wheels. A lot of riders use rims initially meant for time trials in road races - the 80 and 90 mm tall rims.
That's great except for one thing - now everyone had to have aero wheels just to be even with the others. This means that any benefit from one wheelset to another, if they're both aero, will be minute. The very coarse benefits of aero wheels, the minutes saved per hour, have now become, between similar aero wheels, mere seconds.
Now the benefits are "fine benefits", minute ones. Got it? Good.
So what's that got to do with aero road frames?
Well, aero road frames, at least ones reasonably well designed, offer a small benefit over non-aero frames. Within the aero frame market though the differences are barely measurable. Velonews found less than 10 g difference in drag, about 10 watts worth, spread across four frames (if my word problem logic skills are correct - 4.3 g spread for the 2nd - 4th frames, 5 g spread from 1st to 2nd frame.) They claim a 20 watt difference from their control bike, a wide tubed Masi with Cosmic wheels.
The benefit seems pretty small when you stick a rider on the bike, a computer on the bars or stem, and, wait for it, waterbottles on the downtube and seat tube. Suddenly you get a whole bunch of things that coarsely affect drag. Yes, a seat tube bottle has been show to be more aero than not having one at all, but if you have a rear wheel cutout and a very aero rear wheel, I think that a bottle won't help much.
Regardless it's beyond dispute that a bottle on the downtube really screws up the aerodynamics. Cervelo, in their top of the line time trial frames, integrated a bottle into the frame shape, essentially proving that sticking a bottle on the seat tube is not the ideal solution for frame aerodynamics.
Therefore, although I detailed all sorts of things I did on my aero frame, they were all "fine" benefits. Minute. Probably not measurable.
Except for the bottles.
I decided that I'd race with no bottles at all. Instead I'd use a CamelBak for that year (2011). I figured that the little hump on my back would help with aerodynamics, filling in the space behind my helmet, and the bare frame would be more efficient slicing through the wind.
I hoped for a somewhat coarse reduction in drag. Not a lot, not the 2 or 3 mph that I see at high speed with aero wheels, but maybe just enough to save me 5 or 10 or 15 watts average in a race.
My "aero" frame. Note no bottles - that was my "coarse" benefit.
The thing I wanted to do was to reduce my average wattage just enough to give me a sprint. All too often I'd waste myself in a race just hanging on for dear life. I'd get to the finish having averaged 190 or 200 watts and have absolutely nothing left.
On the other hand if I finish the race with an average wattage under 180 I've usually done quite well for myself. Under 170 and I probably did really well.
My goal was to bring the 190 watt races into the 180 watt range. This would allow me to contest the sprint and get a place.
Pause for a bit to catch your breath.
(Jeopardy theme song)
Okay, let's keep going.
Today I realized something. All this is "fine" benefit stuff. The bottle idea was a more significant fine benefit but it was still a fine benefit. I realized this because of a couple things.
First, I've noticed a lot, and I mean a lot of racers who look like they're on a 3 speed bike when they're in a race. They're up in the air, they're on the hoods, they look like they're rolling down the boardwalk to pick up the Sunday paper. I almost always hold my tongue when I see this because I was told a long time ago not to offer advice but to wait for someone to ask for it. People who spontaneously gave advice usually didn't get received well since the advisee wasn't necessarily ready to hear said advice. The problem with the 3 speed position is that it's really upright and catches a lot of wind. It's not aero, and it's way worse than having a non-aero frame. You can't buy your way out of an inefficient position, not with wheels or a frame or anything. It's just bad.
Second, the riders I fit have done well. I have never had the opportunity to fit someone using a powermeter but I did fit a few riders using a trainer to make sure I was in the realm of reality. The racers I fit showed immediate improvement. My fits tend to be the same kind of thing. I couldn't describe it technically but I always found I did the same thing for each rider - extend them, drop the bars, usually raise and move the saddle forward. Basically they went from looking like a rider going to pick up the Sunday paper to a rider that looked like a racer.
I knew each rider would be better but I couldn't tell you why.
Now I can, for a couple reasons.
First, I read somewhere that Michele Ferrari told Armstrong to raise his saddle 2 mm, in order to flatten his back a bit. I thought, okay, come on, 2 mm, that's nothing, that won't do a thing for your back.
Then I remembered that when I put the 170 mm cranks on my bike and raise my saddle 5 mm, it feels like a new bike. I'm much lower in front, my back is that much more comfortable, and I can roll along just a touch easier. I never really quantified it so I just dismissed it as a minor benefit to 170s (and it was a reason I tried sticking with the 170s for a year, then for a summer, before finally giving up and going back to the 175s. The position was great but the long crankarms were an even more coarse benefit.)
Second I had a chance to sort of quantify why the 170s felt better. I saw this site today, for the first time. Don't mind the 1995 graphics. What's important is the link to "Real Customers, Real Results + Improvements". Take a look at what the riders look like before and after. Before - they look like what I described above - a rider on a 3 speed.
Okay, maybe not that bad, but look at any random picture of you (meaning you, the reader) in some random ride/race this year. Look at the picture with a discerning eye. Look at your back, look at its angle toward the lower back area. See how your hips are sitting on the bike.
Now look at the "after" pictures on that site. See how low their torsos are? See how their arms extend forward? See how their hips are angled forward on the saddle a bit?
They look pro now. It's not just a look though. Look at the numbers, how much wattage they're saving, how much faster they can go. It's incredible, the changes. Even an experienced racer saw a huge improvement in performance. Body positioning gives you a tremendous benefit - it is the biggest of the coarse benefits.
Finally, I saw some article somewhere (forgive me for not linking, I've followed so many links recently I have no idea where this was) where someone gave the benefits of a time trial frame, helmet, and wheels. The main benefit came from the frame but not because of the frame itself but because of the position.
This is where all this starts to come together. I know that I know how to fit a rider, within reason. I couldn't even describe what I was trying to do, I just did it. Now I know that I fit a rider in order to tilt their hips forward, let them open up their hip angle, get their back more flat, drop their shoulders and arms, and get their head tucked in a bit more. The pictures from that wind tunnel site really illustrate the kind of results I would go for when I fit someone.
Next, the wind tunnel site quantified for me the savings on the position. They even comment somewhere that aero helmets are aero helmets, probably like aero wheels are like aero wheels. Get something in the right family and they're really close to one another.
At the same time the site illustrated to me that aero frames aren't where it's at. Aero frames may help incrementally but without a good rider position the frame won't be significant. In fact it's probably more important to get aero wheels, maybe some other aero stuff - flattened bar tops, even more aero shoes - before the aero frame becomes important.
So it's position. And position, for most of us, is pretty easy to adjust.
For me, not so much. I have limiters - I have short legs so my preference for 175s means I end up with my saddle back just a touch more than I prefer. I end up with my hips tilted up just a touch more than I prefer. And therefore my back isn't quite as flat as I like.
(I should point out I have a bad back and the flat back really feels comfortable. The other week I got on the bike specifically to help my back, it was really hurting for a few days. The ride really helped it feel better, and I was on the drops virtually the whole time. Another thing for people to open their minds to - the idea that a flat back could be more comfortable than one that's elevated a bit.)
I have been thinking of a third Tsunami, primarily to explore the use of compact drop bars. I really like the FSA Compacts but they're 3 cm shorter in reach and 2 cm shorter in drop. To fit me properly I'd need a 15 cm stem (not happening) which drops 2 cm (especially like that). A longer frame by say 2 cm (and a 13 cm stem instead of a 12 cm, to keep weight over the front wheel) would be good.
The drop presents a problem. The headtube is already as short as possible. With the Chris King InSet headset I could drop the stack height just a bit but I'd want to drop it about 2.5 cm, 2 cm for the bars and at least 0.5 cm for the cranks (since I like it when the saddle is 5 mm higher). The only way to get this extra drop is to raise the rest of the bike, i.e. the bottom bracket. That's not ideal but it would definitely work. A track stem would give me about 1 cm drop.
It's definitely just a gleam in my eye but the idea with this third evolution frame would be to hone body position. It's not as important to have an aero frame (especially for a frame with a 9.5 cm headtube) but having my body position adjusted a bit would really help.
Because on my bike now even I look a bit like I'm out for a stroll to pick up the Sunday paper.