I’ve tried to make Musescore work because they do have the minimum of early music features. But in practice those are still fairly buggy and I gave up.
Horses for courses. If the kind of music you work with fits the feature set of Musescore, that’s great. The people who pay $ or $$$ for notation software include those who either need or want professional output, need to meet specific publisher style guides, need tools e.g. to help with collapsing an orchestral score into a piano version, and/or are doing something other than mainstream music.
Music notation is an ocean of complexities. Different instruments, places, eras, genres, can need different notations. I don’t know of any notation software that can straightforwardly set full mensural notation, let alone play it. It could be done, but there aren’t enough of us who want it for that feature to get further than a to do list. But the more professional software has enough engraving flexibility to fake much of the appearance so that humans, especially students and amateurs, can play it more easily than from either a facsimile or a translation into modern notation. Which brings up a second can of ragworms–playing what’s entered. Human players of many genres don’t play what’s written as written, so each supported style has to be included into the playback software, else it all sounds somewhere from mechanical to wrong even you’re using the sound more for error finding than for production. So many features can start trying to conflict with each other, so it takes lots of debugging. Then it all has to be documented. Unsurprisingly, a lot of that is boring, and above some level of complexity it only gets done well enough if you pay the people who do it.