That's more about why "sphere" is a misnomer that leads people to think the idea requires a large rigid shell than a problem with the idea. You can create configurations of matter that continuously and completely englobe a star without it being a rigid sphere. What you can't do is make it so that any other configuration is always receiving sunlight at all times because no part of it ever in shadow, but that is not a requirement to capture all energy from a star, that's merely an engineering question around efficiency.

The result may not be literally a rigid sphere as people think of it, but if the orbits are packed tightly together enough it'll look plenty "sphere-like" from the next star system over.

A rigid Dyson sphere is probably impossible, and its utility is not entirely clear. Making a Dyson swarm around a star is merely a lot of work, and also has the advantage of immediately paying benefits as soon as the very first satellite is launched.