I would argue that if the comercial software authors manage to convince someone to pay money for it, it means that someone is extracting some value which allows them to pay for said software. It also means that someone is investing time to learn, to install, and finally to use that product.
It's not just about the products themselves (although some products are probably also just clones of something that already existed with better marketing) but rather about everything a software company does in general. Think websites, random general-purpose code and development processes. I think there's a lot of repeated effort in those areas that could be avoided if we all magically knew about each other's problems and worked on common solutions. (Note: I don't think that will ever really happen.) Publicly available frameworks and libraries are sort of that although there's still a lot of "grunt work" overall.
My point wasn't that commercial software is actually useless though. More like, in the grand scheme of things, even something that's just quirky for the sake of being quirky is kinda useful.
