Warning, this may be considered controversial, read at your discretion.
Having read both sides of the argument: to bin or not to bin, I fall on the side that favours oversampling. Risky perhaps when seeing conditions are not very good, however, I believe there are many situations where it would be extremely beneficial.
There are too many interesting objects (galaxies) which are just not large enough to resolve details unless you go sub one arc pixel. Often they may be bright enough to require shorter exposure times so that enough acceptable frames can be used from a large number of captures taken. Choosing to do so puts the onus of responsibility entirely on us, and we need to accept there may be no decent images.
Thinking of something like the enormous Planewave instrument as the one to have this option. Agreed there would need to be a second set of calibration frames, so it's important to only allow this option on one of the scopes where binning 2x2 is the norm. To avoid any need to alter the current booking form, allocate a new pier number to it, where what you enter will be the same as before, but the Roboscopes techies will magically switch the 2x2 binning off for the duration.
I checked and Software tools actually recommends 0.33" to 1.00" / pixel scale under good 1-2" FWHM conditions for such a scope and sensor pixel size combination and warns of under sampling above 1.00" / pixel scale.
Something for future consideration maybe and also something I'd really look forward to using.
Postscript.
Took so long to type in I forgot to include a relevant point. I'd only use the unbinned option when capturing luminance, be that the L or Ha filters. I'd setup a second capture sequence with the default binned option when using the other filters to shorten the total exposure times needed.