No. We're talking about two different factors, that can both co-exist. You were talking about recency in sample size, which I agree with. I was on the other hand talking about development paths.
Yes, the last year is the most important. That's true. But if the last year is fairly equal between two players, and the same for the middle year, will it really matter if someone was a replacement level player in the NHL that first year, or if he was a star in a lesser league? My point is that no, it won't.