Creatine, for instance, has been studied for mental effects a bunch of times - the answers all seem to line up on "can definitely help some people, but probably insignificant to most". Other substances shake out similarly - the data on bacopa is inconclusive, but the studies are numerous enough to suggest that the inconclusiveness is about a very weak or situational effect, not a lack of literature.
The vegetarian point does touch on my larger concern, though: people talk like studies finding weak 'average effects' mean the drug is universally effective, but low-powered. That's possible, but it seems more likely that there are lots of drugs which are situationally powerful (e.g. based on diet or genetics) and hardly any that are good for everyone. Creatine is obvious, and I kind of suspect cholinergics and bacopa for this also. Gwern is really good about acknowledging that his studies (of n=1) aren't being designed to apply to anyone else, but most people don't offer that caveat.
For example, in the article found at the link, note that they only examined 45 people, all of whom were vegetarians (who are much more likely to be creatine deficient due to their diets). Although the subjects were consuming more creatine than the vast majority of meat eaters would consume naturally, since the authors didn't experiment with other doses (e.g. 1-2g a day, more accurately mirroring the diet of those that consistently eat meat) there's no telling whether consuming extra creatine will actually benefit your cognitive abilities if you are already regularly eating meat.