I completely agree. Having a pure-research team with a mandate of "all your research must be geared towards totally reimagining the entire product" is a dumb idea. Having more granular (and collaboratively driven) goals from "some of your research must be geared towards totally rewriting an area of our application that is a major pain point, and for which all previous attempts to do incremental changes have failed for technical reasons" to "look into better tools or strategies we could use to tune performance of, or write better tests for, swaths of existing code" is more realistic and more useful.

Obviously, other, not-pure-research devs should be given time to do some of that work as well, otherwise the research team becomes the "saviors that are always about to come back over the hill" for every other team while they kick their respective cans down the road.