Thanks for the comprehensive feedback!
I'll take it into account for the sequel I'm making.
No problem! I'm sure there's some more things I could extract that make puzzles easier/difficult, but it would take a broader exposure to more puzzles to work out (and the current setup makes that tricky).
I did eventually figure out the sparking, but it took deliberately forcing that one puzzle into different states (specifically ones not related to solving it). I would suggest changing the visuals for the failure state from a radial burst to an arc that causes lightning to jump from the one electrified piece to the other electrified piece indicating that the failure was due to the relationship between the two.
To that, though I'm sure you're aware and take this into account, the different sorts of piece modifiers like that add to puzzle difficulty. And it may be worth investing in creating a hand-designed tutorial puzzle for each of those modifier mechanics and making the algorithm use it at the first point that the modifier would be used. I'm not sure what a tutorial looks like for this particular modifier, as its going to require multiple opening moves whereby BOTH electrified pieces can be moved to touch each other, and both moves look like a path to a solution. Hmm...
Maybe something like this:
The yellow needs to be replaced with a metal square at the destination point of move 2 (as depicted it'll fall into the water; I drew the yellow before realizing this) so that it supports the large green after move 3. Move 3 needs to be made in order to get enough distance across with move 4 (and no other move with the large piece is possible), but move 1 needs to be made to get the brown out of the way first. ANY other move with the brown would result in electrocution or disconnection.