I'm currently on holiday, logging in only to keep the PI ticking over, and use the occassional wormhole to sell product in Jita. By the time I'm back from holiday, I should have the money I need to try setting up my own wormhole, something I'll post more on in due course.
As PI is my primary income stream, I have a strong incentive to think critically about my PI setup, and how it could be improved.
All my PI skills are at IV, and I'm running each of my planets with 3 advanced factories and 6 basic factories. This is the standard setup for single world P2 production. From what I can see on the forums, even with perfect skills this does not change much - the extra CPU and PG simply lets you add more extractor heads to take better advantage of poorer planets.
This approach is very easy to manage, but there are problems. On small, rich worlds (which use very little PG for links), my skills already allow me to harvest more P0 than I can use, and I actually place my extractor heads inefficiently to prevent myself from running out of storage. But on large worlds the extra link costs reduce the extractor heads to below what I need to keep up with my factories. Getting my skills to V would fix this, but it's a long train for a marginal benefit.
I'm also forced to harvest resources that can be converted to P2 on a single planet, and those may not be the best resources that planet has to offer.
So, I'm wondering about a P1 & factory planet approach. The idea is that I harvest a single resource on each planet and convert it to P1 before shipping it to a factory planet for conversion to P2-3.
This is MASSIVELY more efficient because the Extractor Control Units are very PG intensive (PG is the limiting factor for resource extraction planets), and you need one for each resource type you want to harvest. By dropping from two resources per planet to one, I can harvest more total P0, even before I take into account the missing P2 factories. Four such planets can then feed into a P2-3 factory planet, which would do no extracting at all.
The question, though, is whether the added efficiency of the P1 planets is enough to outweigh the loss of production of the fifth planet (which extracts nothing). This is something that ultimately I will need to try. When I do, I'll report back with the results.