Do I understand correctly that you're checking CPU usage immediately after having scrolled up/down, or while scrolling up/down, right?
Apparently, there are already people working on that bug. If I understand things correctly, Firefox uses transparent windows, but transparent windows use lots of CPU on some macOS configurations, and this somehow wasn't detected during testing (I imagine that the computers used for testing didn't have these configurations).
If this is the same bug, as is likely, the bug is identified and developers are working on it: https://bugzilla.mozilla.org/show_bug.cgi?id=1429522 . If you wish, you may leave a message on that bug mentioning that you are available for profiling and testing a fix.
It is my impression (but don't take my word on it) that the fix will ship in Firefox 65, which is currently in Nightly. Now, I'm not sure that the fix has already landed in the current Nightlies, but do you have the possibility of testing on Firefox Nightly if the problem is still present?
(usual caveat: Nightly has Telemetry activated by default – if you don't like Telemetry, don't forget to disable it)