You have to spend some time developing your sandbox. Look at it this way. You seem like you're about to build a prototype with a different Arm Cortex M0, because you can't find a dev kit for the one you're interested in, and that you "can make the application work with as few as 3 UARTs". For this particular step, ask yourself if you really need to make it "work", or do you need to convince yourself that it will work when you move over to the chip you want, and what you need to know before you do the move.
My own preference is to build a sandbox with really strong tools, with the intention of using that to cut down later. You didn't let us know what all your requirements are, which makes guiding you a bit difficult. My experience is that code is fairly portable between ST ARM families, and even if it isn't, redeveloping with a slightly different library isn't a killer. My inclination for a project like this would be to go with the biggest beefiest M0 in the family, and see how big my code is, how fast it runs, how much current it uses, etc, and then pare down-- but again, I don't really know the questions you need to ask this iteration.