Yep. I was doing some experimental designs a page or two back.
I wonder how often that's really faster than just making the items that lead to your desired end product(s) though. And it certainly sounds like it'd be much more wasteful. (Which people argue doesn't matter as much with SA, but I'm coming at it from the perspective of someone playing overhaul mods that don't give so many resources so casually.)
The part I've heard about that gets really janky is making things in foundries, then recycling them to get back parts that you didn't actually use. I think that might be something I'd eschew because it feels cheaty even if it is vanilla behavior. (And it makes me wonder how something like the mod set in my current game, where there's like 5 recipes for low density structures, is handled...)
-Morgan.
So, the problem is that factory complexity goes up rapidly with that situation.
Let me detail the issue on Fulgora itself first:
Scrap is a resource that you can get at quality by putting it into miners, and it
must go through recycler machines that can also have quality to get the base resources of the world.
Two of these base resources are ice, and the planet's unique ore. Both of which are basically useless for quality production because the only thing they can make is a liquid, and liquids don't have quality.
What liquid making processes do have, however, is a need to have all input resources the same quality, and the ore to liquid ore process takes stone too.
So just to handle having quality at the first two steps of the Fulgora resource chain use quality, you need one chemical plant for every quality level, and to keep on hand stone of every quality level.
This is mostly a space problem, stone is one of the base resources so you should get it in large enough quantities as long as you don't waste too much of any given quality on unneeded processes.
Next, there is the issue of lower quality unwanted parts. Let me talk about the mess that is making a high quality processor with maximizing the quality of the end result:
First off, you need to work out if you are using ore or direct casting for the resources involved. Going from ore technically gets you more chances of a quality increase, but with the downside that you need to use electric furnaces for one of those increase attempts, and they only have two module slots compared to a foundry's four.
Then you hit the problem that most later products take a lopsided set of increases. A processor unit takes a lot of basic circuits, only a couple advanced circuits, and a liquid that does not impact quality. Advanced circuits need basic circuits, wire, and plastic all of the same quality. Basic circuits need iron plate and wire.
This means you need to make a very large amount of quality wire, a solid amount of quality plastic, and a fair amount of quality iron plate.
However, each step adds a place where a quality setup can in theory upgrade the quality to the next level, which you want to maximize highest quality outputs, but needs a bit more infrastructure to handle happening.
Here is an example build just to work on uncommon and rare quality module 2s:
Steam Community: Factorio. Vulcanus Rare Quality Module Build
steamcommunity.com
I had to put in multiple "shove the excess into chests" bits here to stop it from clogging up on one or another resource rolling quality too many times compared to other products.
The wire was also shifted from direct casting to made out of plates to help there as well.
I never updated it with productivity modules this build also would want to ensure maximum Rare results.
This entire thing was needed just to make one final product at two different qualities, and honestly I have scrapped most of the uncommon modules this made on Fulgora to get a few more rares using had fed scrapping tests instead of trying to build more arrays like this with them.
You probably could get a large amount of uncommon stuff if you didn't go for scrapping that is somewhat useful, but I'm currently stuck between either upgrading to get Epic at the cost of even more complexity to builds, or just going for Rare with everything using recyclers at key points to maximize rare outputs.
