As far as I know not in a central place, but almost every popular home computer has extensive test suites which test emulator behaviour down to the clock cycle against how the real hardware behaves.
For instance the Wolfgang Lorenz test suite for the C64 is almost legendary and quite hard to completely pass (this reproduces the behaviour of Wolfgang Lorenz's C64, but not necessarily of other C64s with different chip revisions):
ZeroGravitas 2021-08-19 12:13:50 +0000 UTC [ - ]
The sqlite one gets mentioned every so often and I remember being very impressed with the Opus audio codec devs testing plans.
Would be happy to read more like this.
flohofwoe 2021-08-19 12:38:38 +0000 UTC [ - ]
For instance the Wolfgang Lorenz test suite for the C64 is almost legendary and quite hard to completely pass (this reproduces the behaviour of Wolfgang Lorenz's C64, but not necessarily of other C64s with different chip revisions):
https://github.com/floooh/chips-test/tree/master/tests/tests...
Most of those tests have been written in the 90's, so they are now historic artefacts themselves which need to be carefully preserved.