I think I get what you mean, but I don't think that you ahve given an entirely correct description of "concepts": Buildings and their placement do not represent concepts on their own in my opinion (also the buildings are all still present after the change), the lerning is rather about the new implications that those new buildings have:
- The player has already learned to use coal-powered steam-engines for electricity, so using water and pipelines should be nothing new.
- Building pumpjacks is straightforward, like a waterpump just on an oil spot.
- Refineries: Here is where the trouble starts: 2 liquid inputs required to be connected to the right input slot. Also several liquid outputs with their respective slots. This should still be doable at a single learning step, but this is my personal assessment only.
- Outputs blocking recipes: Big trouble! I believe that this is the core problem. Failing to realize that a "full" output capacity is the reason the entire production is stalled may even get players confused about the right way to connect their pipelines: Many will probably re-check every step they made building their oil-setup without realizing the actual cause of the problem.
- Chem plants should again be easy to run as the liquid-input - liquid-output scenario has already been practised with the refinery before.
If I had to make a design decision, I would probably have a gas-only recipe but with an animation that shows the discarded outputs being burned and with extra pollution added. I am not sure, however if such a visual clue could be easily added to the existing graphics. Also there should be a tutorial message pop-up to hint about blocking outputs.
I would also have tow distict triple-output recipes like it is now, as two such recipes in different relations make the process of output-balancing more flexible.
We now have the conversions from heavy to light and from light to gas via chem plant. However, if one needs more heavy in the first place, there is no inverse conversion (and cannt be for logical reasons) from light to heavy. So the option to switch between two triple-output refinery-recipes is actually very useful in shifting the distribution of the 3 oil assets more in the direction of heavy.
Also, as I posted before the challenge of automizing such a switching-process by activation refineries with one kind of recipe and deactivation the others is a nice and welcome challenge.