Definition editing gadget
AFAICT, the definition-editing gadget has not had problems, but I don't use it so I don't know for sure. Has it had problems? How many users have been using it? Has it been thoroughly tested?
I ask all this because if it has been tested AND you think it would be desirable as a default for all or all registered users (Is that even possible?), I would support a test to see whether it would encourage enough additional good contributions to offset whatever increase in bad contributions it might also bring. I'd be willing to do patrolling for the test period.
The tool was tested quite a lot, although most of that was a long enough time ago that it's possible that new problems may have popped up as a result of the many changes to Mediawiki since then.
There was a vote on enabling the tool for all users a while back (Wiktionary:Votes/2012-10/Enabling Definition Editing Options), which failed to pass. Enabling it by default for all users or specifically all registered users is indeed possible, but I think that excluding IPs misses most of the point.
Conducting a test sounds like an interesting idea, but it might be hard to gather enough data to determine whether/how much the tool encourages additional good contributions. Using the Tag filter/edit filter to tag edits done with the tool (so that the recent changes can be filtered to show specifically those changes and such), might be helpful. I'm not sure if that's possible/feasible, though.
I actually think that the key part of the test is whether bad contributions are encouraged. More total edits and a "small" increase in vandalism/bad edits would show success without requiring careful measurement.