Nice question! Research (both clinical and anthropological) plus the anecdotes of others make a useful starting point for personal experimentation, and that's about all you can ask for. Our lifetimes are limited and the combinations of variables to be tried are close to unlimited, so without a starting hypothesis, you'll just flail around randomly. Ya might get lucky, but ya probably won't.
When I started out on this journey, I had a loose plan with some loose justifications (low-carb, since Atkins had worked for me years earlier, plus the carb-insulin hypothesis, plus a sprinkle of Paleo 1.0 thinking). Over time I pushed harder on the actions that got good results, and learned through painful experience to avoid the actions that got bad results. And when research or anthropological observation seems to contradict my experience, I think about why that might be, but as far as my actions are concerned, I end up going with my experience (although I'm careful not to push it onto others, who may differ from me in key ways).
The safe starches debate is a perfect illustration of this: I've tried that experiment -- twice -- and it didn't work for me. I agree that amylase + evolutionary theorizing make a seemingly convincing argument in favor of dietary starch, but I'm not going to try endless variations (eat them plain, eat them buttered, eat Old World tubers, eat them on the new moon, hanging upside down, when the tide is out, and so on) just to satisfy the theory, no matter how well-supported it ends up being in the literature. My body is happy without starch, and unhappy with it, and if I have to try out fourteen different versions of it, each one lasting several months while I gain weight and my joints get stiffer, just to find the one golden version that "allows" me to eat a potato -- well, I hope you can see the return on investment is not exactly compelling.
ETA: This question reminds me of this other question. Also, it allows me to post a link to my favorite comedian doing my favorite routine. ;D