I guess I would reply by emphasizing two things:
1) most of the bend or shape of the power-duration curve occurs/is determined by what happens in the first few seconds to minutes of exercise. So, in posing your question it is important to define precisely what you mean by changes in the individual's personal power profile (as you now have...it was the allusion to "CP" values that threw me off a bit before).
2) relative to the above, the rate of decline in power once you get beyond the first, say, hour of exercise is undoubtly changeable, but to a much lesser extent than many believe/only via prolonged periods of rigorous endurance training. This makes it possible to set up training levels based on a single "anchor point" method, with that anchor point (functional threshold power) being a surrogate marker for power at lactate threshold. At what Kirk Willet would call the "left hand side of the curve", OTOH, things are quite changable, depending on the type of training that you do - however, given the extreme degree of specialization you find in, well, specialists (e.g., match sprinters, pursuiters), it is difficult for somebody to change their basic "type" enough to be truly competitive in an event to which they are initially poorly suited. (Of course, the great thing about bike racing is that people of many different shapes, sizes, and abilities can be successful in one way or another, vs. something like distance running where you have to be quite lean/light to be highly competitive.) In this context, "crossing over" from the first two columns to the second two columns of the power profiling tables is going to be the most difficult, because the physiological factors favoring a high level of performance at the extremes tend to be exactly the opposite. OTOH, somebody who has a high, say, 5 s power will likely be able to train to have a high 1 min power as well (if they don't already), just like somebody who has a high 5 min power will, with enough/the right kind of training, be able to generate a high power at their functional threshold intensity.
Hopefully that helps place things in context a bit better...of course, a more direct answer would be to simply show examples of changes in athlete's power profiles as a result of changes in training focus. I've collected a fair number of those and usually show some when I give talks, but it's not easy to show them here (email me offlist if you'd really like to see a couple).