This is a fine distinction; and I'm not convinced by your selectionism. It is at least plausible that (hormonal?) changes favored because they produced one selectively advantageous pedomorphic trait could as a side-effect have induced another which was neutral or even somewhat disadvantageous.
It is plausible for that to happen, yes, almost all genes have multiple effects. However, the differing timing of the various paedomorphic traits means that there has been no overall neotony of the human form but rather multiple incidences of paedomorphy.
Hairlessness could be non-adaptive, but I find it unlikely. If hairlessness was a neutral trait we'd expect it to be more randomly distributed through populations.