My fines point is simple: using the same grinder, removing the fines on some samples, and not others, then filtering the resulting brews, will yield indistinguishable cups, even with high fines producing espresso grinders. This is easy to confirm over and over in blind taste tests.
Obviously, some people dislike the mouth feel of fines in the cup. But they will probably prefer a paper filtered FP brew to a fines screened, unfiltered FP brew.
The issue of brew grinder testing apart from fines is not so simple.
Different grinders taste different brewed, and are always distinguishable in triangle tests. In the past, this has fooled me into announcing one grinder as better than another for brewing. However, the further tests are never consistent, and have more to do with subtly unequal grind settings and extraction levels, which will favor one grinder with some coffees, and the other grinder with different coffees.
This observation is borne out by espresso grinder tests too. The best tasting espresso grinders are the ones that produce the most forgiving and best flows, regardless of other particle distribution characteristics. But this is precisely the grinder characteristic that is not at issue in other forms of brewing.
Here's a mental experiment. We have one ubergrinder that produces a uniform, goosestepping grind size. And we have one slacker grinder that produces a wide, dithering particle distribution. Which will brew better coffee? Ubergrinder would march to victory every time if one could brew the coffee to a precisely set, optimal extraction level in real time. There the grinder with the tighter tolerance would do better in each round. But if the precise extraction level is unknown or uncontrolled, slacker grinder with the wide distribution has a 50/50 chance of winning any taste test. All real world brewing has this unfortunate uncertainty in actual and desired extraction levels. And this is the reason my ubergrinders have never convincingly beaten my regular grinders in brewing tests.
Finally, there's a reason so few people do and publish blind tests. Nobody really wants to put their latest piece of very expensive gear, and their justification for buying it, on the line.
Of course, this reasoning won't stop people from buying expensive brewing grinders. It will just motivate them to design even more expensive real time extraction controllers