The "version rating" feature that's been put in Wikimedia software but not yet optimized to the point where it can be used in a production Wikipedia would be ideal for reducing backsliding without limiting the ability to edit or otherwise making things difficult. I've been looking forward to this feature for quite some time myself. If the problem is simply one of not having developers with the right combination of free time, interest, and skills to complete this feature, might it perhaps be a reasonable idea to spend some of the money raised in the fund drives to provide a reward to get it done? Spending money to improve the software's scalability could be worth it for the reduced cost of hardware upgrades and the reduced "cost" in terms of the volunteer effort of editors keeping articles clean.