I think cell writes have more to do with changing V folders and re-programming, but i suppose hit counts could be a problem as well. I disabled my buddy's hit counts for this reason. That being said, my Pro-197 has been running 24\7 for nearly a decade with hit counts enabled. I purchased it just prior to Whistlers takeover so it may have a little bit better EEPROM. I have considered buying the WS-1065 as a backup in case it fails.
It's really hard to say without opening the things up and looking at the chip number(s). You very well could be correct in that they improved that chips specs at a later date. I can't say I've really seen any issues related to the failure problems I had in a long time.
Mine were both bought soon after the release of each which was well before Whistler entered the picture so they very well could have an older chip that had poorer specs.
I also ran them with the hit counter enabled so I could see those stats when I read them into Win500.
When the failures started (the 197 started failing first) I did some reading and found where others had the same issue that seemed to be related to the eeprom chip. I then read the specs on the chip and confirmed they had a fairly low amount of writes per cell. This was confirmed by Don Starr as well. I guess that thread is archived here somewhere from way back when! Most of our chat was done in one of the public forums I think.
I do forget if just the working scanlists were stored in the failing eeprom but something makes me think that was the case and that there was another memory chip that stored all the V folder data as well as firmware and the bootloader. Luckily, the eeprom chip that failed is the smaller of the two eeprom chips with only 8 pins so it was not really that hard to swap out.
I also recall Don telling me that every time I loaded my active scanset back to the radio, it was writing to the same eeprom chip that held the hit counts and current scanset parameters.