Hm. So, I think I've reached a point on maxbin2/maxbin-rs where I don't see anything to immediately further improve, so I think I'll let it sit where it is until someone else tries it. It's annoying I have to look at it so abstractly... "On tested datasets, same input, same output, faster, slightly nicer interface, should be easier to install, but Bowtie2 is completely dominating the runtime and the C++ looks... intimidatingly optimized" - so probably no idea if what I've done here is actually useful from a bioinformatics PoV. I really lack the experience to be able to tell if this is good or not.

Thanks so much @Werner Kroneman, I will test it ASAP to check if behaves as the original!

It should but that was just in my tests of course