I think the thing to be careful of before jumping on it too much is what the article is talking about for its comparison. It is looking at number of draft picks/average position picks and then NFL performance. This is not that OSU or whomever has not had loads of picks but how many of those picks are starting level talent over that period of time.
We think of Elliot or Bosa or Thomas but they are not really the norm either. There are many OSU players, or players from any school, that are not cut out to be full time starters in the NFL. It is just how it is. And having more guys like that, which is possible if more of your guys are picked, would hurt the number in this case.
It does not mean that OSU sucks or anything like that. As teams like Clemson and Alabama are negative in the scale too. I imagine the reason is they put lots of players in the draft, they go high and only a small number of them are starters for any period of time. By these metrics, if I understand them right, you are punished by having more players drafted (at least if they underperform). Over performance could mean that a smaller school had like two draft picks but one turned out to be a starter for quite a while.
I am not big on ESPN either, but I do not think they should be executed for this figure at all. It is trying to estimate some things that are useful. It would be neat to see something like this for recruiting purposes, but much much harder to do.
And it really just depends on your definition of an NFL factory. As nonproductive, as you mention, has more to do with general player performance than number taken and percieved NFL talent.