There is a government mandated test for the ratings, so at least in this regard it is a good basis. The test consists of testing the candidate tire against an SRTT (Standard Reference Test Tire), and the tires are swapped back and forth between two vehicles.
But because of the size thing that you mentioned (the SRTT is a medium size), the rating has to be "traceable" to the test in some way. Put another way, if the tire to be tested is way bigger than the SRTT (or way smaller), then you can't find a vehicle that will work.
But frankly, this wear rating thing is just for comparisons, and trying to make too much of it is just wasting time. If you compare 2 tires, both the same size, then you've got some valid comparison.
Put another way, most tire wear occurs in cornering, so the more time you spend in the city, the worse your wear will be. This by far is the largest factor in determining what your ultimate mileage will be.
I've seen the cheapest of the cheap tires get 100,000 miles just because they drove 99% on the interstate.
This also applies to alignment. Too much toe-in is just like driving the tire sideways - not good for wear.