Though I disagree with Logan about the change failure rate (see https://two-wrongs.com/you-can-reliably-measure-change-failu...), I still think he is spot on in terms of MTTR.
Strongly seconding Reinertsen for anyone interested in this, though. I think this article of yours is particularly insightful and of high quality: https://lucasfcosta.com/2022/08/31/engineering-metrics.html