My advice isn’t about this specific layout but about designing layouts in general:
Make a physical test as soon as possible. Start with just paper printouts. Then for the next step, pull all the keycaps off a $5 rubber dome and stick them into modeling clay or similar. If you have access to a laser cutter, try cutting plates that you can stick real keyswitches into. Etc.
Iterate 10 times on your design over the course of a day or two, and then solicit external feedback. Try your prototypes (whether paper or fancier) on your friends and family. Make a few versions and hand them to someone and see what they like and don’t.
Once you have a design that seems to work, build the cheapest working prototype you can. Use cardboard or acrylic or cheap wood for the structure, and try to build something that takes as little design work (no PCBs, no fancy CAD work for the case, ...) and as short a construction time as possible to construct. Hand wire it and use some existing firmware, and don’t worry if some of the solder joints are crappy and break after 2 days of use.
Having something physical to test will (a) help you learn a lot faster than abstract thinking, at least at first, and (b) give you something to show off, which helps keep excitement up both for you and for whoever you’re showing it to.
I see lots of people design fancy dream keyboards without nearly enough iteration in the prototyping phase.