Why would those numbers be relevant? Legit question on my part, I’m not making the connection of how that advanced stat would give any indication of one players impact to the game.
I wrote more above, but in general, the more goals a player is on the ice for, the more likely he's going to get a point on those goals, and with dramatic differences, this can often lead to what's seen as career seasons/down years for players. In general, players have a lot less control over their on-ice shooting percentage, because it's more dependant on the 4 other skaters on the ice, as well as the performance of their opponent, and while elite players tend to be higher than average, it's shown to not be sustainable at extreme highs. Currently Matthews is at almost 13%, whereas the league average is around 7-8. Crosby for example, having a poor year by his standards is at 6.1%, and if you look at spike years for a number of players, it almost always coincides with a high on-ice shooting percentage. It's currently the case with MacKinnon, was the case with Kane in his Art Ross year and was the case with the Sedin's in their Art Ross wins.
Now, that's not necessarily to say that it's pure "luck" driven. I tend to believe it's a sign of the player and his line playing extremely well. You're bound to have years where things just click with you and your teammates and you're "feeling it" and everything is going in. And some years you seem to not be on the same page and nothing seems to be working. As such, it wouldn't necessarily be wrong to say that Matthews has been better than Barkov at 5v5 this year. He's produced those points and been on the ice for those goals and a very high GF%. But they do point to him not being able to maintain that production, and falling back to around where Barkov has been this year going forward, unless he and/or his line improves.