My only point in this dicussion was, is there a "better" system using measurements instead of race time? And can we not take times even when a single fleet is racing against itself, like at a Nationals, where the top skippers are racing? Then that time can be compared to it's P number to see if it needs adjusting. And to throw one more measure into the system, could we measure the distance from A to C mark, so that when we have several fleets racing against themselves, like the upcoming Perf. N/A's, we can gather some data on speed over distance.

Or could they at least keep the time of the fist boat, start to finish, in each fleet, to see how they compare to each other?

As some here have said, getting good input data seems to be the problem with P number racing, as Jack pointed out above there can be other issues too.