I've also seen the work done with pines. Most of the examples shown look really good but I have often wondered how many others don't turn out so good and end up on the scrap heap so have resisted the temptation to try.
A couple of years ago I wired some very small Japanese maple seedlings. To get the bends I wanted for shohin sized trees I needed to start with very thin, flexible trees and used thin copper wire. A lot of them snapped while bending but that's another story. The successful trees were planted in a grow box to get good growth. When I went back to remove the wire they had thickened more than anticipated. I managed to remove the wire from some reasonably easily. Some damage on others because the wires were partly embedded and a few had completely swallowed the wires at the base of the trunks. Just to see what happened I've kept growing all these trees. The ones with embedded wires have swelled way more than the others. As the wires are low on the trunks this has resulted in massive low thickening. As far as scarring, there's still a faint spiral line on the trunk after 2 years but as I still haven't started to develop branches there's still plenty of time for that to fade.
Some initial observations:
Leaving the wire in produces way more thickening than scarring but removing the wires.
Wires must be applied right from root level. Trunks only expand where there's wire so starting slightly above roots means massive reverse taper right near the base. Maybe layering could be used to solve the problem but that's just adding more complexity. Best to wire bare rooted trunks so wires can be applied right from root level.
No point leaving wire on a long length of trunk if taper is the aim. Just leave wire on a short section, length depending on the planned ultimate height of the trunk.
Technique appears to work on deciduous species as well as pines.
After having actually had experience with embedded wire I'm more inclined to believe it could have possibilities.