Turbolifts changing locations is no big deal: it's like changing the location of a specific control panel or a transformer cabinet. That's inherent in the nature of the turbolift, which does not require massive infrastructure, just some holes in the walls and ceilings to fly through. Sure, relocating some of those holes may involve lots of work, but nowhere comparable to relocating something as inflexible as an elevator shaft in today's buildings. (Unless turboshafts for some reason are used as load-bearing structures, or important power or air or coolant trunks or whatever.)
The interiors are just cosmetics. What changes radically is the shape and size of every single hull component: the saucer, the cigar, the neck (plus of course the pylons and the engines). That's not cosmetics, that's plastic surgery. Yet how deep into the body it goes is unclear. The saucer could just have been extended by adding all-new rim structures, and the top and bottom vertices sawed off so that all-new superstructures could be mounted there. But apparently the designers felt the need to affect the curvature of the upper bulge, too, which would be a job much bigger than the rest of it combined. Would some sort of an internal skeleton survive the process? In the saucer, perhaps - but the secondary hull is shown to be hollow, devoid of an internal skeleton. The loads are carried by something close to the outer hull, and necessarily this something from TOS was then completely discarded, because the outer hull no longer is where it used to be!
The inability to hit the right intermix from the get-go just tells us Scotty was right about the engines needing a test flight. It doesn't tell us whether all engines need that, or just the first ones off the production line. For all we know, fifty-two Constitution class ships had had their engines modernized and properly tested and their very personal intermix formulas fine-tuned until ST:TMP came along and Scotty had to cope with a truncated test consisting of just a couple of hours insystem at fractional warp.
If, OTOH, engines straight off the production lines could establish a pattern for the entire fleet, why wouldn't the testing have indeed been completed straight off the production lines, before the engines were even delivered to the San Francisco yards? Apparently, there is a great deal of intergration effort involved.
It's pretty much a case of zero conclusive evidence and complete personal freedom of interpretation whether NCC-1701 was the first or the last ship to be refitted that way. We missed out on most of Starfleet not only between TOS and TMP, but also during TOS. Ships like the ST:TMP one could have been common back during "Where No Man Has Gone Before", or just a glimmer in the eye of a designer during "Turnabout Intruder". The E-A, with its blatant TOS style GNDN piping and "less altered" shuttlebay, may have been a much older refit than the E-nil. Etc.
Timo Saloniemi