A response to a very interesting question:
It sounds like you are asking if it's possible to group together all of the best individually tested aero components and wheels on the best aero frameset and end up with a slow bike. This certainly can be true. The problem with a good amount of aero testing these days is that often critical items are tested in isolation--in other words, a fork can test extremely aero when mounted in a stand-alone fixture, but not perform so well when actually mounted on a frame. The same can be true for different aerobar designs, stems, wheels, etc.
What we're starting to see is an integrated systems approach to bicycle aerodynamics in which component design is factored into frameset design (examples include Trek's Speed Concept stem and front brake, same for Specialized's Shiv, Cervelo P4's bottle design, and on). From a purely aerodynamic perspective, this trend makes sense, as one can optimize component and frame design to work together. On the other hand, this trend also locks the consumer into a limited ecology of aftermarket parts, which either don't fit on these new generation frames or which work aerodynamically poor when installed. The other problem is that integrated designs can work extremely well if the rider happens to fit its specific coordinates, which frequently have limited capacity for variation. (For instance, ever see a P3 or P4, a frame designed with a low headtube, with a huge stack of spacers?)
This past season, I was very surprised to find that the stock fork that came with my TT frame performed much better than it did with a super-aero, 7:1 aspect fork that always has tested extremely fast. When looking at the data, it quickly became apparent that the frame's stock fork was optimized for the frame's unique design features. The take-away is relatively simple: one needs to regard component and wheel selection as part of a more comprehensive analysis that factors in how specific parts influence overall performance.