The model is never certain until you fit it to the data.
That is what's missing. Foundries like to put out couple-
page PDF reports and hold the data real tight. Someone
looking too carefully, might find errors and mess up the
whole qual "done-ness".
"Aging" in the HCE/NBTI aspect is very bias-point-dependent.
Digital qual results are not going to give you what you need.
That is, a pretty fine grained but broad survey of drift vs
conditions, which you would then assess for "no prob" and
"oh noes" regions, and the zones in between where you
have to anaylze rather than adopt / avoid based on rules.
You can include the effects. But to make it worthwhile
and not a false confidence / false panic exercise, you
need to know those effects and what activates /
modulates them.