The concern is not so much about number of parts replaced ($) but about turnaround time because cadence is ($$$). If building a new booster from scratch with it

The concern is not so much about number of parts replaced ($) but about turnaround time because cadence is ($$$). If building a new booster from scratch with its tanks, its engines, the avionics, etc takes n days and refurbishing a booster takes n-1 days, refurbishment's the way to go. More simply put, they could not have hired and built factories enough to keep up with their manifest. I think at peak production they were churning out a bit better than 1 Merlin per day and that's far lower today. So, let's say they could double that peak rate with brute force hiring and factory expansion. That could get them 800 Merlins produced per year. But that's only 80 launches a year at best. They're gonna do very close to 100 this year and well over 100 next year. Further, they couldn't magically double Merlin production as I imagined above. Since boosters are being re-used so often, however, they've actually cut production to something like 1 Merlin every 3 days which gets them a few new boosters each year to replace expended ones, and all the second stages they need.