Your question seems to be about whether the engineering effort required to build a Dyson sphere (or swarm) might be more usefully spent building something else instead.
Unfortunately, that "something else" is not well defined. If you want to harvest power from natural sources, then your only real choices are stars and stellar remnants. The best way to extract the maximum amount of power from most of those is to wrap them up in some suitable layer of stuff. The specific kind of "stuff" in each case may be different, but the underlying principle is basically the same.
There are more energetic reactions than mere fusion, of course... antimatter is the obvious one. But antimatter doesn't naturally occur in sufficient quantity to compete with stars, so you need to harvest stellar energy to produce your antimatter. Even with the most enthusiastically theoretical physics, you're not going to get more energy out of your antimatter than you put in to creating/harvesting it in the first place.
If you open up the possibility of handwavium fringe physics, then all bets are off, of course. But with a grounding in reality, there isn't really much else you could build instead.
There are also some things that are Dyson-sphere sized, and have Dyson-sphere-sized power requirements. One of those is a Nicoll-Dyson laser: you fit the outside of each element of your Dyson swarm (or the entire outside of your sphere, if you've worked out how to make such a thing) with optical phased arrays, turning the whole shebang into a colossal laser suitable for driving fleets of lightsail starships and providing pretty much the ultimate in plausible-physics-based defense for your star system. There's not really any more plausible or sensible way to power this kind of thing than a Dyson swarm/sphere.
Would a Type 2 Kardashev Civilization really build a Dyson Sphere (around their own sun)
. Although the OP of that question intended to emphasize the "own sun" part, they still asked the "would ... really build" part, which is what this question asks. I disagree that the other question is too broad - the eight questions are not distinct, but iterations of the same question "is this idea feasible". $\endgroup$