If you look at the relevant meta-analyses, the effect size estimates are quite modest, so much that most of nicotine's cognitive benefits won't be detectable by introspection.
Small to medium-sized, yes (I assume you're referring to Heishman et al 2010). But modest compared to what? There aren't too many cheap legal stimulants with clean side-effect profiles we can use.
Oh, I agree. I meant modest compared to effects I expect to be detectable by casual introspection.
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.