The problem is that even though the source and the sink are connected with a fluid pipe, there is no fluid transfer. Here, have some more pictures
It seems to have something to do with the branching, even though I switched to block filters and added tank extenders to prevent slushing.
The first two pictures show how the pipe content of hydrogen drops from 200 L to 100 L at the first filter, which does not accept hydrogen. Then it drops from 100 L to 0 L at the next filter, though not in the same block relative to the filter (hydrogen is supposed to flow right to left in the first two pictures and left to right in the third one). So one block to the left in the third picture has 100 L H2. And nothing gets to the third filter, which filters hydrogen. Also, the hydrogen content is static, there is no global or local flow in any pipe. Another thing to note is that hydrogen is inserted into the pipe with a tank extender and it can flow in two directions. The first I just described, and the second is a dead end with no hydrogen filter. There are two other filters though, but the content is 200 L anyway. It might not have anything to do with filters, it’s just an observation. Fact is, though, that the content is static and goes from 200 to 100 to 0 before entering the hydrogen tank.
I tried to disconnect the pipe in a few places and a little bit of hydrogen went into the tank (just emptied a few pipes). When I reconnected the pipe, the point where it goes from 100 to 0 moved closer to the tank (it went past the 2nd filter). Still no flow though.
Did it again and H2 started flowing into the tank. Then I disconnected the pipe at the tank extender that inserts the hydrogen and connected it again after about a minute. Same situation as in EDIT2, except it goes from 200 to 176 to 100. Actually… I didn’t check that it EDIT2, might have been the same. Anyway, no flow. If I fuck with it a little it starts flowing, I doubt that will last long.
Now N2 and SO2 stopped flowing from the centrifuges. O2 seems to be going, still.