I'd say that's actually a good-sized sample. I mean, let's be honest, if this study showed that his PPG didn't change at all, it would be used as evidence in his favour, not brushed off as a small sample.
1.35 PPG is still awesome, but it's not otherworldly. Sakic did this well over an 8-year period and the league was much lower scoring (and he was good defensively). Yzerman topped that by a fair margin. Clarke's best 8-year period saw him average 1.22 PPG while being infinitely more valuable at the other end of the rink. Trottier averaged 1.42 in his best 8-year period which should make him a slam dunk better player considering his own intangibles. There's a big difference in where Espo's theoretical ranking is, if we start to imagine he's scoring that much less. It goes from "almost Lemieux-like" to quite beatable.
This represents an 18% decline in production and I think that's the bare minimum he's looking at.
Of course, as you alluded to, we're just talking about "raw" numbers and the deeper numbers are less flattering. I've fiddled with the 1968-1975 numbers for hours and with Orr having an R-on of 2.19 and R-off of 1.10 and Espo 1.64 and 1.41 over those years, it's clear that they didn't actually spend a heck of a lot of ES time on the ice together (or their numbers would mirror eachother a lot more). The best estimate I can come up with is that the team was 2.28 with both on the ice, 2.15 with only Orr, 1.22 with only Espo, and 1.02 with neither.
Not that any of this matters, since he's getting voted in. I voted him in myself, and I probably cast the 2nd or 3rd-lowest vote for him.