I would recommend you try to determine the real foundry
used by MOSIS when that flow was actively supported,
and try to find archived copies of that PDK. Maybe at
NCSU, Idaho, or whoever else was an early MOSIS adopter.
And of course the authors themselves might be willing
to point you at some larger body of work, or even share
obsolete design files if they're academics with no "IP
envy".
But that aside, I think you could just stick with the
topology and use whatever foundry info you deem
relevant - only voltages, length effects and speeds
ought to change much.
The other alternative is to dig around for medium
voltage power management processes; something
in the 12V range is likely to be a ~ 2um gate length
and a 300-500A gate oxide.