@francisfordpoopola
It seems a bit sloppy to run water where water isn’t needed just to control it. I’ve not ruled it out entirely but I think that implies cutting two ¾" holes to run ½" pipe in each direction, which is perhaps structurally comparable to removing a whole brick. And if I remove a whole brick, I might as well just stick my hand through to reach the valve. Though to be clear I don’t know if it’s one layer of brick or two. Since it’s interior, probably one.
Regarding PEX, I have very little confidence in it. I just fixed a huge amount of damage because PEX that was only 13 years old leaked enough to collapse a ceiling. I also have a PEX compression fitting that I had to replace every 6 months routinely until I decided to replace the fitting with a longer run of PEX. Normally I would favor soldered copper. However, the input side of the valve is steel (copper-incompatible) and the other side of it is PEX. So indeed it’d be steel (with many threaded joints) or PEX with fewer joints. At least ½ the PEX joints would be accessible.
@Gordon
angle grinder: The gear ratio is probably around 4:1 or more so you’d either be turning the knob forever or it would be very difficult to turn
Once or twice a year, I wouldn’t be too bothered with 4 rotations for every valve rotation. A ball valve is only ¼ of a rotation anyway. Although the valve I’m favoring ATM has a replaceable cartridge which is not a ball valve but I think it only needs 1 or 2 rotations. If I needed to close the valve often, then I would consider the inverse: making it hard to turn but then put a big wheel on the wall & do a steampunk style.
Regarding the wood gears, the pegs had not occured to me and perhaps they are a good approach. Quite simple and may even work with scrap Ikea particle board. I know there are some youtube vids on right angle gears in wood but I’ve not watched them yet. I know flat gears can be done with just a drill and jigsaw (though I don’t have a jigsaw).