Custom keyboards took off because of mechanical switches. Back in the day people wanted mechanical switches because they last longer than membrane ones, and so you wound up with a bunch of companies producing relatively easy to manufacture mechanical switches. Those switches all felt and sounded a little different so you got people who wanted a specific feel and sound and it grew from there.
There hasn’t really been the same push with mice because even really cheap ones work really well. Optical sensors are way harder to produce than key switches, and while there are a few different ones on the market other than dpi and polling rate they kind of all act the same - it kind of either tracks right or it doesn’t. There’s no differentiation unlike switches that are “tactile” or “linear” or “scratchy”. And because of size restrictions you can’t really have the same kind of switches as keyboards use for the buttons. And unlike the really niche keyboard people who do their own PCB and machine their own case, making a good mouse on your own from scratch is way more difficult. They’re weird shaped and it’s much more difficult to change things like optical tracking algorithms compared to macros on a 40% keyboard. You can do a run of 100 super niche keyboards and make it work, but just the injection molds for one mouse mean you need to make 10000, which stops it being a project and makes it a business.
There are premium mice manufacturers, but in general they either are going super light, super ergonomic, or super functional - and honestly they have a hard time competing with a company like Logitech that can produce really similar features for a fraction of the cost and have a decent reputation to boot.