I haven’t tried the Blackwell method. Of course I think brms is great but the me implementation is very top heavy and demanding with big data. Something that needs more work. Also not a lot of theory on what prior to use for measurement error (ie normal or something more flexible).