Amerigo, I don't completely understand your question. CHIRP is not a subsisitute for RMS (Winlink) software. Instead, CHIRP replaces or augments the Yaesu programming software for for your radio. You can possibly (probably) run your RMS Windows software on your Linux Mint system by running it under Wine, and you can almost certainly run the Yaesu software under Wine.
If you are asking about support for the FT-991 within CHIRP, then sorry, it's not there. CHIRP is Open Source software written entirely by unpaid volunteers who like to program and who like Ham radio. Each radio type needs to be separately analysed and a driver program written for it. The volunteer will need physical access to a radio to perform the analysis, and will need an absolute minimum of (say) 10 hours of work. A volunteer who can do this in 10 hours can probably make more than $200/hr on a paid contract, so clearly this must be a "labor of love". Mere mortals will take longer than 10 hours. My guess is that in general a volunteer might program a CHIRP driver for a radio that they have chosen to use themselves, but are less likely to program for some random new radio unless there is a change that the new radio turns out to just need a minor modification or adaptation of an existing driver.
Speaking personally, I am a new CHIRP developer (and a new Ham). I purchased a Yeasu FT-4 before I had even heard of CHIRP, and was unhappy because I later discovered that CHIRP would be highly desirable in my application but the radio was unsupported. So, I wrote a CHIRP driver for it, and incidentally also for the FT-65, since they are very, very similar. It's been a whole lot of fun, but it took multiple weeks of effort. From a purely financial perspective, it would have made far more sense to just buy a supported radio or even go out and get a 1-week programming gig and then go buy an even more expensive radio.
If you have any programming experience, Please consider trying to write a driver. It's really quite fun. Even if you don't have programming experience, you might want to try to do an initial analysis if the protocol, using some variation of the techniques described at :