FYI for add-on developers:
If you use ComputedAttribute in your add-ons -- and you plan to upgrade to Plone 5.1b4+ (which incorporates some changes by pinning ExtentionClass 4.3.0) -- you may want to look at the wrapping level you declare (related to acquisition). I used a method decorator as described here, and ended up needing to change the level I declared from level=1 to level=0 to avoid exceptions due to ComputedAttribute not returning what was previously working.
Grepping my omelette in 5.1b4 install, it may well be possible that there are (possibly untested?) corners in core packages using wrapping level=1 via something like ComputedAttribute(fn, 1) that might or might not work. I am not sure if my case is isolated, but it is worth leaving a note about this here.
@seanupton thanks for reporting your findings!
Would it be too much hassle for you to report it also on Products.CMFPlone issue tracker?
The grep output would be useful there, so we can add tests to them and double check that they do work.