Re: ether v THF
Everything is relative. THF can become deprotonated to give ethylene and the enolate of acetaldehyde with butyllithium at room temperature. Grignards, less so, but.... THF has also been used to isolate R2Mg compounds (presumably because MgBr2 precipitates?). Peroxide formation?
Re: addition v insertion
Interesting question. This may be a lot more difficult to KNOW. I seem to recall iodo-esters reacting by insertion as a faster reaction, or something like that. Since halogens are known to react as electron acceptors, e.g. Br2 + alkenes or R-I + I(-)
RH + I2, one might guess a similar reaction can occur with Grignard formation or a transmetalation. If that were the case, the difference in reactivity may be greatly influenced by the R-group. Even if a Grignard were to add/donate electrons to the Br, if no further reaction took place, it would be a "no reaction" reaction. That is, donation of electrons to Br2 can give an R-Br-Br](-) intermediate, but will readily lose Br(-). That is much less likely with an alkyl bromide.
Since I like to argue that reactions generally follow similar pathways, I like to seek a similar pathway to ones we know. We know Br2 reacts with Br(-) to give Br3(-). Then transmetalation reactions may occur by an initial addition to the halogen. This may be how arylhalides are transmetalated. I'm not saying an addition/complexation must occur, but that a series of reactions is known that makes the question a reasonable one to think about.