Well, by the book, the system should just perform as if Cat 5 components were installed everywhere. The UTPs are all the same nominal impedence, so things like impedence lumps and standing waves shouldn't be an issue (assuming all components are properly installed)...or at least they shouldn't be any more of an issue than a straight Cat5 installation.
SOME Cat5e / Cat6 cabling use 24/25ga wire instead of the usual 26ga...so it's conceivable (that being the case) that the punches are light/flakey (like a dinner roll?? no...as in marginal).
The biggest difference in performance is going to be from the jumpers to the panels (stranded into solid core). Most certification happens panel-to-panel, so perhaps fabricating some solid-core jumpers (for the purposes of testing) might be worth a try.
Ultimately, after you've looked at the signalling with a Sniffer-like device (looking for late collisions, and framing errors), and found some problems, the only course of action that'd give you any legal standing would be to hire a third party to come in and run certification testing on the plant.
I'd think a walk through the contract would be worth a shot too. If the contract is for Cat5e everywhere, make 'em change out the panels and cross-connects. If the contract is for Cat5 everywhere, I'd try to get 'em to choke up a certificaction report or bring 'em in for random testing of ports you designate. If X% of the ports fail, they repull or re-terminate untill it passes.
I'll try to talk to the folks at the Anixter Lab and see what they say about it. I'm in class this week, but I should be able to squeeze it in on a break.
Kind of a goofy situation, I hope you do a better job of hiding the body (of the person that arrainged this install) than the last time....
FWIW
Scott