A useful question to ask here is whether there are characteristic differences between the sorts of projects that are distributed freely even in an environment that supports commercial distribution, and the sorts of projects that are distributed commercially when that option is available. If there are, then it's similarly useful to ask whether the second category has any value above and beyond what the first group has.

For example, if I compare free fiction to purchasable fiction, do I find that the latter group has valuable properties relative to the former group? What about if I compare free software to purchasable software?

If not, then your argument makes sense; there's no value being added by commercial projects, so there's no cost to eliminating the system of constrained distribution on which commercial software projects depend.

If so, then it's useful to ask whether that extra value will still be available if that system is dismantled.

Personally, I often find commercially distributed projects that offer signfiicant value relative to freely distributed ones, which is why I often pay for commercially distributed projects such as books and software. Presumably, if I didn't find that to be true, I would be perfectly content to use freeware/shareware products instead (as I do for computer games, for example).