1.So the 'Rule of 105%' is a pretty old school type rule of thumb. I conceived it in the early 2000's when we were experimenting with V shaped, parabolic, hybrid toroidal and early toroidal rims.. basically we found that they all would sort of work as long rim was at least 105% of the tire width measured almost anywhere on a deep rim. Granted there are differences in effectiveness of this which are the pillars upon which the various wheel companies have built their respective empires, but really, the rule holds from a very binary good/bad sort of perspective.
That is to say that there is more or less a near zero change of your wheel/tire system being fast if the wheel and tire are of equal width.. it really only works if the rim is wider. (and really only works if the rim is at least 105% of measured tire width)
This makes sense if you think about it from the point of view of the air... the tire is the leading and trailing edge of the wheel system and by the nature of it being built on an inflated casing..it really can't be aerodynamically optimized... so you need a rim shape that collects the separated air and smooths it out when you are looking at the front half of the wheel and a shape that delivers the air onto the tire in a way that it it most likely to 'close the gap' when you are looking at the rear of the wheel.. and to do either of these, much less both of them, the rim has to be a good bit wider than the tire.
As far as where the rim is widest is a sort of shell game...you can do it all at the brake track which makes the front half of the wheel very efficient in both drag and lift...but then the rear half of the wheel is less efficient and you have very forward centers of pressure and potentially unstable handling in cross-winds. You can move it to the middle of the rim and get pretty neutral handling but potentially at the cost of higher drag as the tire gets larger..or you can move it more toward the spoke bed which can do some cool things for handling, but can make some tire shapes no longer work well... so much that even tires within the rule of 105% don't work so well.. so we seem to be at a point where there are tradeoffs between total drag and handling, but for the average person were are talking pretty small differences in both.
2. So my point about the asymptotic relationship of width to rolling efficiency was mostly meant to point out that the 'wider is better' mantra isn't exactly true if you are considering pavement. As the returns are diminishing with width increase, there is very little benefit of say going from 30mm to 32mm and even less going from 32-34mm...and on smooth pavements, the difference is often immeasurable. The point here really being that there is some real benefit in rolling resistance and comfort moving from 23mm to 25mm and slightly less, but still measurable moving from 25 to 28mm but as you keep widening, the curve is so flat that the differences appear to be minimal to non-existent.. whereas the costs of getting there become ever higher. So for something like a Tri bike, the actual benefit (if any) of moving from a 28 to a 30 or 32mm tire is so small that it's almost certain to be wiped out by all of the other compromises that have to be made to get there...in particular, there just aren't really any aero rims that will work all that well with 30-32mm tires.
Now if you begin to consider other factors and surface conditions this will change. Our data looks primarily at various roughnesses of pavement and the Crr curves relative to tire width are pretty flat by the time you get to 28-32mm of tire width. However, if you move onto dirt, or gravel, or cobbles, the curves change and you find that wider tires suddenly have huge advantages again. This is where a bike like the new 3T Strada is pretty interesting and very fun (full disclosure, I am good friends with those guys so my opinion is likely colored by that!). Here is a bike that is built as an aero road bike that can handle tires at what seems to be the max width for both Crr and Aero when considering paved and related surfaces.. And when the surfaces move to gravel or dirt or worse, you move to something like and Exploro or an OPEN UP which can handle 40+mm tires which are smoking fast on the rough stuff.
3. Personally, I ride 28mm Corsa G+ on Zipp 303 rims which measure 29.8mm at 62-64psi and fit nicely inside Campy Calipers. This is a nice blend of comfort at some aero expense for our bad roads here in Indiana, but more importantly it adds some comfort to my daily driver which is a 1993 Team Issue Eddy Merckx MX Leader (Molteni paint scheme for those wondering). This bike is completely irrational, way too heavy, too stiff, terribly un-aero, and too overbuilt for most anybody ever (mine was originally built for the Eddy Merckx Podio team) but at the same time, makes me smile to look at and also very happy when I ride it. For me, the wider tires at lower pressures are of great benefit on our terribly rough roads full of pavement seams and such as they offer excellent handling and comfort (plus latex tubes.. oh so nice!)
If I were racing or at all serious about going fast, I'd likely generally be on 26-28mm tires (measured) on one of the very aero bikes out there for most events (likely run at 75-90psi depending on surface) and in the right conditions or the right events with high quality pavements I'd happily move to something like the Vittoria Corsa Speed TLR at 23mm (will measure 24.8-25.0 on wider bead seat rim which means I'd likely run 85-95psi depending on surface quality) for the combined Crr and aero benefits. The real key is having options and knowing when/where to use them and having wider rims and frames that can accommodate, certainly allow for more options.
http://www.SILCA.cc Check out my podcast, inside stories from more than 20 years of product and tech innovation from inside the Pro Peloton and Pro Triathlon worlds!
http://www.marginalgainspodcast.cc