OpenMSX ported as a libretro core, is there willingness to do so?

That indeed used to be the case, fortunately though it has come some way towards a more proper implementation for computer platforms. Computer keyboards are supported nowadays (thank god )

For example, current BlueMSX libretro core supports full keyboard. It can be seen in the file libretro.c in the Libretro BlueMSX git here: BlueMSX libretro. Another good example is the Commodore 64 core (VICE-Libretro), which has full keyboard support also and works very well too.

For a better experience I would suggest to make use of the "enable hotkeys" option. It's in config -> input -> hotkey mapping -> Input hotkey binds and "enable hotkeys". This allows for the emulated machine's keyboard to be active unless the hotkey is pressed (which allows Retroarch GUI hotkeys to become active).

You're right though when you say it's still not fully perfect. It has its roots in providing the best setup for consoles, and it still shows, even though it has improved over the years. The biggest issue which you touched upon is that without workarounds you can only load into the primary rom slot or floppy. I understand providing direct access for subslots is in the cards, but currently there's a workaround needed for loading multi floppy games. The Commodore 64 core for example allows "cmd" files to be loaded, which are just plain text files that have a file path in them for each floppy in drive 1 to N.

Another interesting core is the Sharp X68000 core (px68k), which both has full keyboard support, but also enables the original emulators' GUI via a hotkey. So I can imagine an OpenMSX core could make use of similar techniques to allow for as much functionality as possible.

In case it's of interest, the documentation on implementing the libretro api can be found in the libretro docs wiki here: https://docs.libretro.com/specs/api/

I believe one dev said his was interested in helping but he wasn't going to spearhead the project. We are still doing this as a hobby

Thanks, that's much appreciated. I'll contact Aliaspider, who was involved initially in the BlueMSX libretro port and ask if he possibly has an interest. I guess he should be able to judge the amount of work involved better than most. He does seem quite busy with other aspects of Libretro though, so apart from interest I'm not sure whether he has the time. Hopefully something fruitful will come out of it.

Could you elaborate?

I'll be willing to help in any way I can, even. But there is no one in our team going to take the lead for that port, it seems.