It's not clear to me that laser links would require reorienting substantial portions of the satellites mass. From a theory standpoint the only thing that would need to be moved is the lens, and even that may not be necessary[1]. While I'm not familiar with the intricacies of point to point satellite communication, I'd venture that it's within physical possibility to maintain fiber like levels of bandwidth over large distances in LEO.

While a major leap, Starlink could potentially service internet backbone traffic at substantially lower latencies in such a world.

1. https://en.wikipedia.org/wiki/Phased-array_optics