> The D120 value is for one *specific* cell. Any cell of course, but
> still the total number of times *one specific cell* can be rewritten.
>
> The number of writes in D124 can be spread over any number
> of cells.

That sounds reasonable

It's worth reading AN537 to get an idea of the factors affecting
EEPROM endurance. In that AN, and others, Microchip say
that although EEPROM can be characterised to a large extent,
enviromental conditions and even data patterns do have an
impact on endurance

Personally I'd take the manual parameters as just a guide and
use the information in the engineering ANs to get the best out
of the EEPROM, or at least prepare for the consequences of
Write operations under less-than-ideal conditions

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.