Test suite needs to be more exhaustive.

I notice that most of the folders in mtest/libmscore only contain a handful of tests, and that in general the tests are limited in scope. They also tend to be just checking of issues that have previously been reported as broken. So when I've added a bug fix, I've adhered to this and only added test cases for the specific thing I think I fixed.

But I think the test suite should much much larger and contain not just behavior that was reported as broken, but should basically cover as much functioning behavior that musescore currently does (within some reasonable limit) to gaurentee that all the currently-functioning behavior will continue to function in to the future, by quickly catching regressions the moment they are sent to github. I also want to note that the test suite when run on travis with -j2 only take 30 seconds (and that is including the longer benchmarks), which is a small amount of time compared to the ~20min that it takes for travis to compile the test suite. I think we shouldn't feel ashamed for utililzing much more of that time...could hypothetically go up to ~25 minutes for tests and still fit within the 50 minute limit Travis imposes (and could hypothetically exceed that with concurrent test jobs). Keep in mind when debugging on home computer, only really need to run the particular module we are debugging, and we can all utilize our own personal travis for running the entire test suite (https://musescore.org/en/developers-handbook/git-workflow#Run-tests-on-…).

There are also some behavior which is harder to fit in the specific folders already setup in mtest/libmscore, so one question is is it ok to create more folders, and is it ok to maybe create a general folder for some things that might deal with behavior covering multiple modules, so I don't waste time trying to decide which folder to put some test in? Also is it ok for me to submit PR's containing nothing more than tests for things that are already working?

Part of the reason I bring this up is because when submit a PR and someone says have you tested it, all I can say is that I've fiddled around with some things that I think might be affected and that of course my special test covering the bug passed, but I can't say with much confidence that I haven't broken some (possibly remotely related) previously working behavior not specified in the test suite, without having a more exhausitive test suite.