Many of these journals require copyright assignment, at which point it's not your own work anymore. Just one more reason the traditional scientific publishing model needs to die a quick death.
Many? More like... all of them! As a scientist, I am fucking sick of copyrights. Maybe they're useful for some (but certainly not all) artists, but for scientists they are nothing but a way for big media (and Elsevier, Nature Publishing etc. are big media) to wrestle control of the scientist's work away from the scientist him/herself.