I realize that it's market pressure that's shaping wheel design to a large extent and I agree with you that the vast majority of triathletes--and TT riders, too--will benefit from the current designs. In a lot of ways, the new designs from Hed and Zipp are honest acknowledgements of the capabilities of most competitors.
For me, I'll be best served by a high performance wheel that excels in low yaw conditions; but, as you point out, the market for such designs is very small. Fortunately, there are some wheels that are nicely designed for low yaw angles and narrow tires--it's just that they are getting scarce these days. As a hedge, I'll be looking to ebay for a couple of fall back spares. But for me, it's so wrong to see a big ol' beefy tire on the front of a TT bike ;)
Tri is where the dollars are--no disputing this.
I've done lots of field testing with my PM--Veloflex Servizio Corse, 20mm. Extremely well-glued on all samples, all samples glued the same fashion five months earlier on the same day (I'm obsessive, okay!), same PSI, etc. Did multiple trials with two different forks (Blackwell Time Bandit and Funda) on mulitple days, on multiple courses. Differences weren't huge, but enough to make me want to use the Easton. Also used my Hed 3 as a baseline.
Difference--what I can say is that is that my H3 is a faster than my Easton or Zipps at very low yaw angles (say, <5*), though between 5 and 10ish the Easton trended faster. Above 12-15 (a really windy day), the Hed 3 just gets faster.
Fork-wheel interaction--still struggling with this one. I've tested faster with the Time Bandit, but its fork blades are very wide and thin--it has a 7:1 profile. My gut sense is that wheels with continuous surfaces, like deep Vs and discs might like forks tight against the wheel, while composite spokes do good with wide-spaced forks.
All of this said, I was seriously impressed by the stock QR fork on my CD 0.1--this fork as a distinct bowed design and is very fast.
Don't know the answer to this one--if there is a difference between the Funda and a UCI legal wider fork (Trek, Oval) and a composite spoked wheel, I don't think that the differences would be bigger than the differences seen in actual wheel design, if this makes any sense. Need to focus on this variable next season.
I ended up switching from my Hed 3 to the Easton 'bout halfway through the season in that it seems to be a good match for the wind/yaw conditions that I mostly experience where I race (some wind, normally courses run on a loop). If I travel any distance, then I just bring the Hed 3, as it's a safe choice, all in all.
Honestly, it's been quite a few years since I've used a wheel shallower than 90mm for TTs; I don't find any handling difference between a Zipp 808, 1080, Blackwell 100, or the Easton; they all seem very predictable and stable. The Hed 3 does handle a little differently, perhaps a bit more unstable in gusty/transient cross winds, but it's pretty easy to control. You can get Eastons at a good price these days; it's remarkably strong, too--I had one on my QR when I got pegged by a truck this past August, and nary a wobble on it. Of course, I wouldn't encourage you to replicate this specific experiment ;)
Really good questions--like I've said, I haven't looked too closely at the fork spacing-wheel design interaction yet. The design differences between the Blackwell fork (very thin, wide space) and the Funda (UCI legal, narrower) are so great that I don't want to extrapolate right now. I didn't see a difference in ordering among the wheels when swapping out the forks; there was some compression of the results when looking at the data associated with the Funda, but not so much that would make me want to say that a wide spaced fork is best for all wheel designs. My personal preference, though, FWIW is to stick with thin, wide-spaced forks as they seem to work extremely well with all sorts of wheels. I plan on trying out an Argos steel fork next season (steel, 4:1 ratio); I wish I still had my old Hooker fork for comparison.
Yes, the Zipps did better than the Hed 3 at mid-range yaws; on the other hand, keep in mind that many people--perhaps most--significantly over-estimate wind speed at ground level. (Rough rule of thumb--wind speed at 30 feet will be 1.5x greater than wind speed at ground level. This is a pretty big difference.) I think that one might be surprised at how low average yaws can be when riding at speed.
I've done some tunnel testing; I use this data as one point of external validation of the trends that I see in field testing, of which I do quite a bit. Yaw angles can be estimated pretty closely by superimposing weather data on a GPS track--have a family member who's a meteorologist, which helps a huge amount.
I tend to go with two wheels--my Hed 3 for windy days, or for racing on courses that shift direction quite a bit, or for when I'm not sure what the day's average yaw might be; my Easton for when I'm pretty confident that I'll be in the yaw sweet spot for most of the race. Really, you're right in that the Hed 3 still is a very fine, fast wheel, despite its design age, and you could spend a lot more money for not a great factor of improvement.
I'm seriously interested in the firecrest wheels, as they seem to address the tire/wheel interface problem common to alloy-rimmed designs. If I only could get one that's 100mm deep and 19mm wide, I'd probably be one happy camper.