I was with the space industry in India. The aviation sector is uniform throughout the world and uses feet, ft/s (vertical rates), knots (air and ground speed) etc. But I believe ground ranges are in kilometres and fuel loads are in kilogram, though I have heard pounds used in some places. Some ex-Soviet countries used to work with SI units even for aviation. But that difference between them and the world was partially responsible for a very tragic and horrific mid-air collision over Charkhi-Dadri near New Delhi in 1996. I don't know if they changed that afterwards. Meanwhile the naval and marine sectors also use nautical miles (different from the imperial miles) and knots exclusively. I believe that it's because the naval conventions were formed before the SI system was devised. Aviation sector just borrowed from them.

Considering all these, you'd expect space sector to borrow from the aviation sector. But we use SI systems exclusively. Everything in metres, kilograms, seconds. Feet, miles, knots etc are unheard of (Well, we have heard of them. We just don't use them). SI units make calculations and our life a magnitude of order easier. I need to check up how it is with winged reentry vehicles. But they're also likely go with m/s rather than knots. The only time we face difficulty with esoteric units are when we use some rare sensors. You end up looking up the definition of 'BTU' and other similar atrocities.

There are two noteworthy exceptions to this trend though. It's when specifying engine thrust and specific impulse. Engine thrust is often specified in kilograms, (metric) tonnes etc. Of course they mean kgf and Tf (weight equivalent of that mass under 1g). Meanwhile mN, N, kN and MN are also used equally frequently. It's a perennial source of frustration and conflict, with younger generation preferring SI units and the seniors preferring kilograms and tonnes. Meanwhile, specific impulse is even weirder. If you were using SI units, you'd expect N.s/kg or m/s or something similar. Even if you were using imperial units, you'd expect something similar. But the unit everyone actually uses is seconds. For examples, a high end cryogenic engine may deliver an Isp in the range of 450s (SSME had a vacuum Isp of 452s). Sometimes, it's also expressed as 'effective velocity' of exhaust in m/s. There are logical explanations for all these weird units. But the reality is that none of them, including the SI units are strictly correct, because they all use some sort of scaling that isn't linear or an assumption that doesn't apply.

You can blame the US for all these inconsistencies in the space sector. The Americans have a habit of making up units on the spot. For example, the kT, MT yields of nukes were invented by the Manhattan project scientists. Similarly, the unit of nuclear criticality is dollars and cents - thanks to Louis Slotin. (Sadly, he passed away soon after the second criticality accident with the demon core). Anyway, the US also has shot themselves in the foot by mixing up units. The Mars Climate Orbiter crashed into the planet instead of entering its orbit due to the engineers mixing up the SI and imperial units. Moral of the story, if you plan to go to space, you better choose a measurement system and stick to it. Also, don't make a round scrubber for the command unit and a square scrubber for the lander. Make up your mind first!!