I'm no physiologist, but I think I understand research in general well enough to make some comments.
Physiology, it appears to me, is a science is where medicine was in the 1970s -- still very concerned with short-term laboratory studies using experimental designs and relatively small numbers. The experimental protocol allows you to say somewhat definitively that the difference in treatment was what caused the difference in outcome.
But the price of experimental design is that you can't generalize very well to other situations. It's not clear how you extrapolate from a controlled six week study of 20 athletes to recommendations for people in the "real world," where you can't control the training (treatment) very closely.
In medicine, and relatedly in sociology and economics, the last 30 years have seen a great deal of innovative work about how to measure the effect of a treatment (in this case, increasing your miles from 70 to 100mpw, say) on outcomes (race performance) in the real world. This kind of observational research doesn't seem to have influenced studies of running performance very much. Will Hewson from Otago University in New Zealand did a long-term study of athletes back in the 1990s, and lo and behold actually did find that increases in mileage continued to have benefit.
Long-term observational studies are difficult, even if you're just trying to vary one factor, because the other factors keep varying and the people being studied (runners, patients) have a tendency to do their own thing and not comply with what they're meant to do.
To bring this back concretely to the question at hand, to answer the question of how much mileage affects performance in competitive runners you'd have to recruit a relatively large number of runners and observe their training, and get some of them to reduce mileage and some to pick it up, by a set amount. You'd also have to do the performance measures in a controlled way, which would start to get expensive bringing in (maybe) 100 athletes for a treadmill test several times over a 6-24 month period.
I'm not saying you couldn't do this kind of long-term real-world study of mileage and performance, randomizing one group of 70mpw (on average) people down to 60mpw and the others up to 80mpw. But it would be a large and complex undertaking. At the end we might learn that, on average, a 10mpw increase in mileage led to a 2% improvement in performance, while a 10mpw decrease lead to a 4% reduction in performance, and that there is a lot of individual variation in response to training. We already basically know that!
Until such time as one could do this study (and who would fund such a thing?), what we know through accumulated anecdote continues to be true. Train as much and as hard as you can given your limitations. When you're pushing close to your physical limitations increase volume and intensity with some caution, because injury will set you back more than the extra 5 mile tempo run will benefit you (or whatever straw breaks your back).