We are getting a lot of test fails based on timers in OSX10.14 which indicates that something may be wrong with the timer. A lot of these tests are being modified with new boundaries to account for the timer problems.

Some of the affected cases:
https://bugzilla.mozilla.org/show_bug.cgi?id=1558893
https://bugzilla.mozilla.org/show_bug.cgi?id=1561410
https://bugzilla.mozilla.org/show_bug.cgi?id=1544452
https://bugzilla.mozilla.org/show_bug.cgi?id=1558882

Is there anything in particular that has changed in 1014 to produce these errors?

Possibly, but nothing in particular that I would be aware of.

Do we know whether this is an issue with timers firing at the wrong times, or with our code executing slower than before so that we need longer timeouts? Both would be concerning.

Ah, bug 1560411 and bug 1544452 comment 7 seem to indicate that it's the former.