.. and 99% of those probably spend 100% of their time trying to sit utterly still in their orbit, staring down at Earth, maybe twitching around to point a bit if they're doing spy work.
JWST pivots the entire spacecraft to point the 6 meter mirror. I'm sure they've optimized the consumables on the schedule, but it simply 'costs resources' to do this sort of work. I assure you there are gruellingly dense and optimized observations plans afoot.
Compare to e.g. the very successful Herschel Space Observatory -- it too went out to L2(-ish), did everything it was meant to do in its three year mission, and managed to extend to four before (drum roll) running out of fuel, this time for cooling. JSWT solves this cooling issue by some incredibly funky mechanical resonance cooling. Herschel "only" cost about $1 billion, and it will be interesting to compare the science results.
Also don't forget it at no point can it turn around and face the earth or sun or the light coming from the suns direction will burn out the optics. If it does it once, and it's mission over. All the while essentially driving most of the way up out of the gravity well and letting it get pulled back to earth a little and repeating. If it goes too far over the peak of the hill it will go over the other side with no way to reverse even if the rocket has the fuel.
I think you are misinterpreting where the fuel consumption comes from. It isn't the reoriantation. That you can do with internal reaction wheels. All you want is rotation, not movement. Turn something inside in one direction and the outside will turn the other way.
The only problem is movement for that you need to eject stuff the opposite way you want to go. And that is needed because L2 is unstable. I.e. you roll down over time. Which needs to be counteracted by station keeping. I'm pretty sure they'll use the fuel usage to measure the local gravitational field around L2 and from that update mass distribution estimates around the inner solar system.
Station keeping, yes, is probably going to need the lion's share, but I have no idea by what factor. I have read that dealing with reaction wheel saturation is also another fuel consumer though. Perfect reaction wheels attached to a spherical cow would use no fuel. As would perfectly balanced solar pressure engineering etc. It basically all ends up nibbling some fuel.
JWST pivots the entire spacecraft to point the 6 meter mirror. I'm sure they've optimized the consumables on the schedule, but it simply 'costs resources' to do this sort of work. I assure you there are gruellingly dense and optimized observations plans afoot.
Compare to e.g. the very successful Herschel Space Observatory -- it too went out to L2(-ish), did everything it was meant to do in its three year mission, and managed to extend to four before (drum roll) running out of fuel, this time for cooling. JSWT solves this cooling issue by some incredibly funky mechanical resonance cooling. Herschel "only" cost about $1 billion, and it will be interesting to compare the science results.