We have looked into drag and solar sails, EM tethers, as well as installing modules to provide retrograde propulsion. We agree though, the best approach right now is just docking and moving at the moment.
The use of combustion based propulsion is doomed IMHO, except for imminent collisions that warrant a rush launch of a mobile, attachable thruster.
My preference would be the perfection of a number of attachable(glue or epoxy or clampable?) ion jets with a suitable steerable solar cell to power it along with a refillable mother ship with a dozen of these deployable systems within it - perhaps in a standard large cube sat that is large enough to boost and refill the mother ship - which would have it's own solar array to power it's own ion jets to meet the prey to be de-orbited. Once there the small de-orbiter would be released close to the prey, whereon it would deploy it's own solar array and steerable ion jet system (which might be a set of 3 for steering. I agree, their thrust would be low - but continuously applied until depleted it would act to lower the prey to a higher drag orbit.
In addition, I might suggest that all satellites be designed with a number of standard push-on attachment fittings, perhaps with orientation ridges(to prevent rotation) alongside so one of these thrusters could attach and lock to the prong without need of glue or threads. I am sure that such an attachment prong can be designed.
There are fairly frequent cube-sat launches that could, for a fee, be willing to orbit more of these thrust recharge cube sats to add to the fleet that is gradually going to solve this issues.