Having spent a fair part of my life working in the criminal justice research field, the biggest blocker I see to something like this is political.

I can only speak to Massachusetts, and I can only speak to the time ~ mid 2000s when Rhode Island and Massachusetts were dealing with claims of racial profiling - specifically driving while black - but there was a great deal of friction with having a central store of any aggregate data on specific judges.

The idea that you could easily run an analysis of a judge's decisions cross referenced by, say, race, was such a complete non-starter in most professional judicial working groups where you'd be looking for buy-in on something like this. Having a good deal of these records available in disparate systems, or even better, only in paper form, was actually a concerted decision at that point in time.

I'm really interested to see what kind of traction you see (and where it comes from). Good luck!