I also found that recruiting endogous Dynein or an overexpressed Dynein motor domain is sufficient to speed up MT-RF. Recruting a motor deficient point mutant had no effect (see video). I saw the same also for the overexpressed minus-end directed motor domain of KIFC1.