I don't blame Wysh for predicting that 12 of last year's top 20 scorers would post higher point totals this season. The old adage "if you don't have anything nice to say, don't say anything at all" applies to player projections as much as it does to anything else: the more players you project for regression, the more anger it will stir up.
But regression being what it is, it's very unlikely that 60% of those guys will improve next season. Here's what's happened over the last two seasons:
The NHL is an extremely competitive place. The reason that players crack the top 20 in scoring in any given year is as much due to good luck - both in scoring and good health - as to any significant skill difference relative to the players ranked #21-#40. Over the last two seasons, only 25% of players in the top 20 in scoring scored more points the next season, and without checking, I'd imagine that the last 40 years don't look too different.
Bottom line: when you're projecting player scoring, always bet the under. And then get ready to duck when people start throwing things at your head.