My gut reaction: "metric used to measure stroke recovery favors bilinguals"

The article mentions cognitive tasks that are likely to benefit on some level from "thinking in language". Duo clearly unrelated tasks like motor control also correlate? Is this really about brain activity keeping the system "fresh" and more capable of repair, or is it maybe just a simple matter of someone with "more words" pre-stroke having "more words" post-stroke?

My wildly uneducated guess, illustrated with computer analogies: when the brain has to cope with multiple languages, "language thinking" overflows from dedicated to general purpose "hardware" and due to some evolutionary dice roll, the dedicated stuff tends to get taken out first in a stroke.

I've only been to India once, so obviously anything I say should be treated as purely anecdotal. In rural Karnataka, I did notice, that although literacy rates (writing/reading) weren't nearly as high among the poorest population, it seemed to be the norm that they knew Kannada along with at least one of Konkani, Tamil, Urdu and/or another local language. English was also very common, but not among the poorest.