The "choice" concept is interesting, but in my view poorly executed.
Instead of "Pick A, B or C and forget the options you didn't pick" I'd change it to
"Pick A, B or C, those you didn't pick get more expensive or it's effectiveness gets reduced."
So the "choice" instead becomes "I choose to specialize and I get a benifit, but I can get the other choices at a cost (time and/or lost opportunity)"
In a way it exists with the weapon techs. If I specialize in beams, I can still research missiles or mass drivers, but resarching all of them may be non-optimal
If you just want to keep the "exclusive choice", just prompt a choice when researching to make the result "unique" something like
"While researching Beam Miniaturization our scientist found a way to improve our original resarch by making beams even smaller, making them cheaper or more effective. Unfortunately picking one option will make the others impossible to research.
Pick Smaller [25% Mass reduction], Pick Cheaper [10% cost reduction, 10% mass reduction], Pick Effective [+10% Range, 10% mass reduction]"
As a side note: we need to bring back the Miniaturization line of research
But one thing you don't want to limit at all: extreme world colonization.
The way it was in GC II, a "basic" colonization (50% effectiveness) and a "non-limited" (100% effectiveness) colonization will work great with the concept of "specialization".
If you specialize in "Barren", you get both techs at once, the other colonization techs in the same tier require more research
Fact: In an old SD game called "The Corporate Machine" this "choice implementation" exists (at least in theory)
eg: if you research"Windows", "OS/2 will get more expensive". if you research "OS/2", "Windows will get more expensive" (Computer product)
I also remember CRT/LCD techs (Computer product) and Jet/Prop Engine techs (Aircraft product) with the same "research one mans extra cost to the other choice"