Regarding circle vs. J hooks, which was mentioned, and Mike gave me a PM regarding it. There's this study in Idaho:
https://collaboration.idfg.idaho.gov/FisheriesTechnicalReports/WildTroutSymposiumXIHigh2014%20Performance%20of%20Circle%20Hooks%20When%20Bait%20Fishing%20for%20Stream-Dwelling%20Trout.pdf
The results are compelling. I'd note a few obersvations about the study:
1. I'm a little uncomfortable with the "active vs. passive" figures. It's not really a criticism but a recognition that it's a variable that is very difficult to control, and it's not fully explained HOW it was controlled.
2. Note that the bait they used was crawlers. From personal experience, crawlers are among the worst baits in terms of deep hooking %. I'd guess deep hooking rates would be cut by more than half if they had used live minnies, or even reduced if they had used eggs, millworms, and the like. Fish just seem to try to swallow garden worms and crawlers more quickly than others.
3. For flies, they used dries from size 4 to 14. From personal experience, that's probably among the BEST flies to use in terms of deep hooking %. I suspect deep hooking rates would more than double if they used, say, nymphs from size 16-24. It's still double a low number, so rates would be low. But you deep hook with dries less than nymphs because you see the strike, even anticipate it sometimes, rather than relying on some indication to show in a slack line. Plus I think the fish going up, grabbing, and going down with the fly tends to lead to some hooksetting by the fish itself compared to say, a fish holding in current and inhaling a passing nymph. Plus, size. Bait, fly, or spinner, bigger offerings go deep less quickly. And for PA, frankly, a size 4 dry is ridiculous, as we consider 12 and 14's "large" for dry flies, and that's the smaller end of the study.
Anyway, those are little nitpicks and I don't think they really confound the conclusion that circle hooks improve deep hooking rates considerably for bait fishermen and may be something to consider in terms of potential regulations if you want to reduce mortality but not exclude the bait crowd. If they confound anything, it's any attempted comparison between bait and fly or active vs. passive. And more than confound, I think it more accurately clarifies that there's just a lot of variables in those comparisons, which makes it difficult to control and capture in any single study.