by Tycho » Thu Mar 18, 2010 3:36 pm
Quick release mechanisms, especially on front forks, tend to "walk". As the dropouts move around due to frame or fork flex the QR skewer stretches and rebounds like a spring and the hub outer locknuts can walk around in the dropout. If you keep your QRs tight and have hubs with knurled outer locknuts, rear wheel ejection shouldn't be a problem, but it is something to keep in mind. On flexy QR disc forks you often find that the caliper will begin to rub now and then, and you have to loosen the QR and re-seat the wheel in the dropout. I haven't found this to be a problem on the rear wheel, probably because frames (especially the Mantra rear triangle) are much more dimensionally stable than telescopic forks.
There are two forces acting on the rear wheel under braking, the force of the caliper or brake pads (and the reaction of the wheel to that force) and the force of the rider's momentum wanting to carry the bike forward (and the reaction of the tire/ground interface to that force).
The wheel will want to move in the opposite direction from which the brake is grabbing it. In the case of the hayes mount caliper location, the wheel would want to travel upwards (or roughly into the dropout). In the case of a higher mount adaptor, the wheel is going to want to travel backwards, or straight out of the dropout.
The natural tendency of the wheel is to want to pull out of the dropout (assuming rear loading dropouts) under braking because the rider inertia wants to carry the frame forward and the rear wheel is opposing the rider momentum, but with the OE caliper location some of the braking force (namely the torsional force reaction) opposes this. However, with a high mount caliper these forces add, which may or may not be problematic as stated above. Seems not to be a problem based on the posts here and on the main page posts, but something to consider.