Messing with the signalling systems can be productive in the short run, and certain individuals have other needs, but it is unreasonable to assume it would be good for most people forever. I do drink coffee every day, because I like it, but I would never assume it makes me more alert or smarter in some way.
I'm not claiming that caffeine will make you super-human but the use of substances like these that may have decreased survivability in pre-historic years (e.g., due to increased metabolism, induced bowel movements that decrease nutrient absorption) may augment several aspects of human cognition (and worsen others) in a way that is towards optimal, for say, coding but suboptimal for, say, creative writing (I don't claim this is true, just provide this as a possible example). Many contemporary college students can definitely attest to both the positive and negative aspects for drugs such as amphetamines (e.g., Adderall) in a manner that one could describe (though misleadingly) as superhuman.
I do agree that use of many of these substances can have seriously deliterious effects with prolonged use, but all studies I have come across refute this possibility for caffeine.
For example I was far less productive in the mornings and after eating, and more alert in the evenings and when I wanted to sleep/relax.
But unfortunately I have relapsed and am drinking a coffee right now...