Well, little ones are easier to prototype. They also are less likely to get in the way of anything. Focusing on effectiveness, though, would probably be better to retrofit a bunch of large ships to drag it around and process the fabric on-board. If the little ones are effective enough once deployed, that will probably be the ultimate outcome. But the technology, right now, is unproven, and so no one is going to invest the millions[1] it would take to retrofit ships.

The distinction between large robots and large manned ships is important to note, though; I was thinking that Kliment was referring to large robots, and therefor didn't even consider manned ships (tunnel vision can be irritating, sometimes). You're thinking about large, manned ships, with no need to function autonomously, based on standard ships, and therefor able to be repaired without as much effort - minimizing the amount of new technology in a solution makes it less likely to break, outside of the new technology. That is obviously superior to a swarm of small robots, except for prototyping.

Also, swarms of small robots are considered shiny technology right now.

[1] I don't know how much it would cost. I'm guessing a lot; take the number with a whole lot of salt.