Oh I get it now. The Space Age foundry is an Ultracube structure. It's supposed to use the cube.

seen from Germany
seen from China
seen from United States
seen from United States

seen from United States

seen from United States
seen from Algeria
seen from United States
seen from United States

seen from Türkiye
seen from United States
seen from Japan
seen from United States
seen from United States
seen from Spain

seen from Türkiye
seen from Germany
seen from China
seen from United States
seen from Türkiye
Oh I get it now. The Space Age foundry is an Ultracube structure. It's supposed to use the cube.
wdyt of ultracube? Appeals to your discerningly narrow taste?
First, my taste is not "discerningly narrow," severe depression just makes you make lots of excuses to hate things.
Second: I quite like ultracube in principle. In practice I find it quite fiddly but mostly for user error reasons. I've never really used combinators for more than very simple contraptions, so I'm not used to thinking in their paradigm. I kept getting really frustrated because I could think of enable-conditions much faster than I could wire them up. I kept thinking please just let me write a line of code and getting too frustrated to enjoy myself.
Also I don't understand what it is with some overhauls (IR also offends here) and starting players with personal construction bots that are blatantly worse than placing stuff manually - it's not like player-placement and bot-placement within the same blueprint naively sum together to a faster construction speed. (I think I may be the one person ever who thinks static roboports are more useful for construction than personal roboports - the biggest advantage of construction bots without high levels of bot speed is they can build where you aren't from an expandable inventory that your mall directly pushes items to. but for some reason every evangelist of rushing bots specifically means rushing personal bots, which are only beneficial over manual construction for a small array of tasks.)
I guess the alternative is you can build this sort of counter to do your ratios on a single machine. But that's still quite fiddly when you want to repurpose it for a different process and fluid throughput is a menace as always.
This is the most efficient way I could find to produce replication gel (at green science, anyway.) Currently it has jammed (output full) because it's a test build and it's just set to loop forever instead of having an activation condition. The bottleneck of this build is the transfer speed of basic matter units, which will be fixed when stack inserters and loaders are available.
Another double edged sword of ultracube: you need to shut down all of your production to test your test builds, which is obnoxious. BUT if any of your builds need manual priming, that's much less of a liability because you're never going to copy-paste most builds that use the cube.