Yes I think I’d agree with that. However, I don’t necessarily believe generalists can’t be as much of a specialist as specialists, even temporarily. It’s just that they are much more rare. Extremely rare. You’re looking for true polymaths.
Part of it I’d contend, is to do with that magic number of 10,000 hours that folk assume they need to have gone through to be a specialist. It’s a number I fundamentally disagree with, as people learn at different rates. Apparently, psychological studies have also found this number to be false, though of course, nobody can learn in zero time.
Another thing I’d argue is important is the familiarity of the concepts that a new thing is based on, as much as the new thing itself. For example, when programming, most imperative programming languages have loops, variables, conditional statements etc. The first language we learn requires us to learn the syntax AND the concept. The second, less so. We automatically sort the concept from the syntax and it is that second language that teases out the concept, making the third, fourth, …, nth language easier to learn. It’s that ability to generalise through that abstract transferability that makes us human. The idea of ‘concept cells’ in the brain, regular neuronal clusters which appear to manage abstract concepts, regardless of their physical manifestation, is gaining some ground last I heard. Though ironically, I’ve not looked into that area recently… don’t laugh.
Concepts is one of the reasons I’m deeply into applied math. Once you have that, nearly everything else, inside or outside tech, can be defined by nonlinear dynamics in that ‘language’. It also makes it really easy to pick up nearly everything in a short time relative to most other noobs, though you won’t of course, be as fast when you start as those who have had a very significant head start, even if you accelerate past them quickly later. Though some find that until it’s mapped into a frame of reference they’re used to, it can be hard to get started.