What I do wonder is if you want to increase chelation control in the reaction why isn't toluene/benzene an appropriate solvent? Butyl ether is a whole different beast compared to THF and diethyl ether as well from what I have heard.
And yeah, the self coupling thing is a royal pain when it can occur. I see it some with the chlorides but we have been able to sidestep that just by dropping the temperature. With the bromide nothing helped though. I never saw formation of the desired grignard reagent using a variety of conditions (dilution, stoichiometric ratios, solvent, temperature, magnesium activation methods). For future note, we found that the best results to avoid self coupling have come by using magnesium powder activated by stirring for 48 hours (at least overnight, but longer the better) under an inert atmosphere (ideally in a glove box). This method will also activate turnings if thats all you have.
My PI hates grignards because of how finicky they can be. Organolithiums, at least in his experience, behave much more predictably. It would also be interesting to know what organolanthanide species like cerium or ytterbium would do in DCM. I would imagine that would be a poor idea in general as they are considerably more nucleophilic than their magnesium or lithium counterparts (while being less basic).