I believe some platforms endeavor to continue incrementing TimeTicks even when coming out of suspend mode.  There are also other, even more frightening behaviors related to suspend mode, like TimeTicks::UnixEpoch() breaks if either TimeTicks stops increment, or the local clock is changed.

We could just flush the cache whenever we enter suspend mode, but that callback isn't available on all OSes (I think it's only available on Windows and Mac, and I think Windows is the platform where TimeTicks increases even when in suspend mode, though I could be wrong).

It may even make sense to implement our own logic to try and fix TimeTicks to account for suspend mode ourselves.  This no doubt affects a lot of other consumers as well.

That is what I was thinking

https://dev.chromium.org/developers/design-documents/sane-time and https://cs.chromium.org/chromium/src/components/network_time/network_time_tracker.h?l=52 are attempts to address a similar problem