The P2B boards were built with good caps, so it's unlikely to be those (even though they are old at this point).
Years ago when I used to get bulk lots of motherboards, I occasionally ran into boards that had this kind of behavior (working only when underclocked). Not sure why it happens.

A worrying thing is that it might continue to degrade until it's not stable at any usable speed. 🙁
I wonder if some aspect of the power circuitry somewhere is failing, like a dying MOSFET or something. The tests you've done don't show any obvious problem with voltages though.
One possibility (no real evidence, just a random guess) is that it could be a function of power draw from the Vcore circuit. If that were the case, then it's possible a CPU that draws less power would be able to work at a faster clock.

I'm not really advocating spending more money on random parts.. but do you happen to have a low end Coppermine? These boards frequently support Coppermine voltages but it depends what regulator chip you have. It's the ~20pin chip between the LPT port and the CPU.

Alternatively - does your P2-450MHz have a partially unlocked multiplier? Some Deschutes chips will let you set the multiplier lower, but not higher. Might be interesting to see if it can run at 100FSB with a lower multiplier like 3X or something. But I think most likely it's multiplier is locked.

I looked through the thread but didn't see whether you actually confirmed the board is getting and distributing power - do the fans (case and CPU cooler) spin when you turn it on?

Both fan connections get power.

And yeah shamino i think my P2 is a locked one. I tried setting the multiplier to a different setting in 66FSB and got nothing it would stay at 300 Mhz. And i turned it to 100 FSB and tried to lower the multiplier to something lower and still no boot. But i was wondering too if its all 100 FSB CPUs or just the higher up ones. I think ive decided against putting it at 75 FSB. I wouldnt want to further hurt it.

I was reading that most 1.00 Revs dont support Coppermine, but i was very interested in popping one in to see if that turned it on. If i run into a very cheap one ill buy it for sure.

It's not a big sample size, but I've seen 4 P2B-F rev1.00 boards that I remember, and at least 3 of those (IIRC it was all 4) had a Coppermine compatible VRM. They just needed a BIOS flash.
If your regulator chip (between the LPT port and CPU slot) is a HIP6019CB, it can't go lower than 1.8V. But if it's a HIP6019BCB, or a US3007CW, then it can, and just needs a BIOS flash.
There can also be other chips there which I don't remember, but they can be looked up.

Well i got my hands on another P2B-F 1.00. Pentium II 450 works just fine. Its using BIOS 1013A. But when i boot into Windows it gives the Error -- Cant Write ESCD. And this was before i Flashed the BIOS and wasn't giving me this error when i first got it either.

Only a fraction of bad capacitors show any visible signs of being bad.
It's very common for Aluminum Electrolytic caps to be bad but look perfect visually.
.
After about 15 years of age the failure rates (even of so called 'good brands') start going up markedly.
.