For the most part CPU dies that are from the inner part of the wafer will be able to clock higher than the ones on the outside of the wafer.

As I understand it, it has a lot to do with tolerances and of course the machines that make the CPU dies are not going to have looser tolerances the further they operate from the center.

AFAIK (over)clock capabilities are related to the silicon/die quality.
You can have two processors made the same day out of the very same "wafer", one barely meeting the minimum specs of the given model and the other capable to run at much higher frequency than stock.
Usually the cores selected from the center of the wafer are the most perfect and capable and are used in the high(er) end models.

It also used to happen that, to satisfy high demand, some very high quality cores (intended for high end models) were used for mid/low end ones, creating overclocking monsters

So I guess that the short answer to your question is: yes, it's pretty much a lottery

Interesting, I'd always thought the P75 was only marked down for market pricing, and would take 100MHz without a sweat.
Even more surprised that the multiplier is unlocked for 2X

I think the 75MHz Pentium was introduced when Intel realised the production yields weren't very good and chips didn't reliably reach 90/100MHz. Instead of scrapping them they were rated for 75MHz and sold at lower price. This is why the 75MHz model was introduced after the 90/100MHz models. Those early P75 chips usually do not overclock well.

Later Intel refined their process and got much better yields but didn't discontinue the P75. Those later models reach 100MHz pretty reliably in my experience.

I haven't had any left sharks. My C2s do 120 but the E0s, man, put on your seatbelt, 133..150... and boot at 166 (83x2) but it was overwhelming the HSF I was using (Not expecting to go past 133ish) and I didn't do serious stability testing.

Doornkaat wrote on 2021-09-10, 21:34:

I think the 75MHz Pentium was introduced when Intel realised the production yields weren't very good and chips didn't reliably reach 90/100MHz.

This doesn't match the story I heard about the P75. I heard the story that the P75 was introduced as laptop processor as a less power hungry Pentium processor. Because low power was more important than performance, the low bus clock of just 50MHz was deemed acceptable. The P75 thus was not introduced due to bad yield, but due to marketing demand (every P90 could be used as "low power" P75), and overclockability can be quite good.