Is this for the purpose of more easily identifying causative factors? Kind of like always developing/debugging in the same environment for reproducible results?
Yes, but at the same time most inbred strains have genetic disorders precisely because they are inbred. E.g. the most commonly used strain C57BL/6 develops age-related hearing loss. To avoid this, there are other lab strains like CD-1, which are deliberately bred to maintain a controlled level of genetic heterogeneity.
Yes. But also I think there is certain amount of selection for mice that will behave and cooperate with lab protocols. Meaning, mice strains, that for instance, aggressively destroy stuff, or sleep too much will not be used.