Good point about the different language groups. I wonder if there has been any research on measuring the scale of the effect by including some sort of distance metric between the languages. Let's imagine your primary language is Spanish: What's the gain if you learn Italian (Romance language) vs German (non-Romance Indo-European) vs Hungarian (non-Indo European same alphabet) vs Japanese (non-Indo European different alphabet)?

And the dot plots suggest no rate decreases until 2028.

No rate decreases until a “real” recession, but yes, early 28 sounds about right.