Mike,
good points, the angles make a big difference for distance races. This is a tough issue...
And here's another thing to consider: fast sailors competing make the boats faster. In other words, the newbie on a P18 may have an easy rating, but he/she doesn't have anyone to tune against, and generally sails with old sails. Also, old boats are slower. If almost all of the P16's in the U.S. are waterlogged and flexible, then this is actually very representative of the "actual" boats being raced.
While it may not be good if Randy, etc. decides to sail a brand-new P18, this rating IS accurate based on the current sails being used and the level of tuning within the class. I don't know... seems to me that it's a tradeoff between accuracy within the system from a mathematical perspective and a sort of "self-handicapping" system.
That said, personally, I would be inclined to support the sort of "ratings can drop, but can't go up" idea suggested.
I don't think we should go back through the portsmouth #'s and "revive" the lowest #'s possible, however. This would be unfair to those in the "dead boat" classes, as a lot of tuning has been lost. It's been a slippery slope, and maybe it's better to just see that it doesn't go further. I think it would be better to just force the classes to perform at at least their "current" levels. Then, we don't slam the newbies with beat-up boats with impossible #'s and lose more sailors.
Ultimately, what it comes down to is this: We're all leery of some great sailor coming along and blowing us away on a "dead boat" because of it's cushy rating. But on the other hand, if this happens on a regular basis, then the rating will be adjusted accordingly.
Frankly, I have a strong hunch that Bill Roberts exploit this weekend was to prove a point, not to win a trophy. More power to him. Hopefully, more people will do what he did. And guess what - the # will drop, and the rating system will self-adjust. I'm not sure what system is better, but I just thought I'd point out some of the advantages.
Just my take on it.

Michael Coffman
t4.9#32