If anything, I'd alter the smaller ones as rotary cannon in space are rather silly.
No, the concept actually makes the same amount of sense as it does in atmosphere, at least if you're going to bring a firearm into space to begin with.
You'd still get the benefit of a lower rate of barrel wear, but, as others have noted, the primary reason rotary cannon exist is for improving cooling, almost entirely via conduction to surrounding air
Gatling weapons being able to fire at a higher ROF without overheating isn't simply an issue of conducting heat to the surrounding air. It's primarily an issue of each barrel in a gatling weapon being used to fire a proportionally lower number of rounds. The reduction in barrel wear and the reduction in overheating are both directly related to the same issue of each barrel in a gatling weapon being used for only a fraction of the total overall ROF. The barrel wear is directly proportional to the heat generated from rounds travelling down the barrel which is in turn directly proportional to the number of rounds fired through that barrel over a given period of time.
which is absent in space.
The heat generated from the operation of a firearm is due to the chemical energy produced by the propellant charge, which is in turn converted to kinetic energy when the bullet accelerates down the barrel, and a proportion of this is in turn converted to thermal energy when the bullet moves down the barrel and heats the metal. As I described above, this would still apply equally to barrels of a gatling weapon operating in a vacuum which would each see a proportionally lower ROF due to the lower individual ROF per barrel.
Non-rotary autocannon would make more sense...
If you want to maximize the rate of fire and were limited primarily by the weapon overheating then a gatling autocannon would make the most sense.
fewer barrels could be made more durable
Which is still less effective than simply using a gatling design.
and non-rotating ones would be easier to attach to a fluid cooling loop.
Although some liquid-cooled autocannons do exist, i.e., the water-cooled 2A7 23 mm autocannons on the ZSU-23-4, a liquid-cooled system is comparatively more complex than simply using a gatling design. When applied specifically to combat in space there would also be the added challenge of designing an effective cooling loop that can operate properly in a wide range of environmental temperatures (i.e., deep space vs. combat near a star) as well as the high g-forces experienced by a rapidly maneuvering spacecraft capable of up to 10 g's of acceleration (given that the liquid coolant will experience those g-forces as well).
There are also several benefits of an externally-powered gatling design in terms of reliability since the action will continue to cycle as long as the gatling mechanism has external power, which means the weapon will continue to operate reliably despite misfires as the misfired round will simply be ejected. In comparison a single-barrel autocannon would be much more limited by a misfire, which would prevent the action from cycling, and it would still not be able to achieve a comparable rate of fire to a gatling weapon.
These issues are presumably why the Expanse, which probably features the most realistic depictions of sci-fi combat using real-world weapons, uses gatling autocannons as defensive weapons to achieve a high ROF against small targets and for point defence against torpedoes. They also look very similar in design to what we currently have in Elite in terms of being a compact, self-contained turret design:
