Well, I finally got around to updating my spreadsheet to include PI. Some interesting things are popping out: Raw materials are worth more than I thought. P1 and P4 are where the isk is. P2 and P3 are spotty. There has been a fundamental paradigm shift in infrastructure to fuel ratios. Now that fuel prices are actually tied to structure prices (since all the fuels area also components in a lot of structures), the ratio of fuel cost per month to structure cost are loosely linked. They used to be fixed. And the ratio has changed.
It's now cheaper than ever before to buy towers and POS modules (for the most part). However it is now much more expensive to run a tower. It's a good thing ice products are so cheap or this really would have hit the fan. It should be noted that this should not affect the small time tower operator in any way. Usually if you're a small time operator you'll be able to setup your main or your alt to do fuels and you'll be able to to keep going much as before. However once the pre-PI stockpiles run out, the major industriallists will need to take into account the new fuel costs for their moon harvesting and moon goo reaction towers.
There is good news: It is now cheaper than ever to replace infrastructure. The bad news: it's now more expensive to run the towers. This will have a long term impact on the price of moon goo, as well as POS research. Tower operators will now have to take the increased fuel costs into account when calculating the cost of their end products. We should see a reduction in proliferation of towers in the long run as the increased cost forces people to normalize their POS situation and make sure each POS they have out there is serving it's purpose in a cost effective manner. It also means that POSes are going to be easier to replace than they were in the past. This should lead to a reduction of the "fear" of pos loss being a limiting factor on POS placement decisions. But should lead to a more "cost centric" restriction on POS deployment as a whole.
One thing I think CCP has messed up in it's concept with PI is that it's mostly safe time wasting for a carebear. This is a nice change of pace compared to most other activities. Mining? constant exposure for the entire operation. Moon and refining? risk your infrastructure. Need to move goods from A to B? run the blockade. With this endeavor (PI) you only need to be exposed to danger during the hauling phase of the operation. And there are ways to minimize your exposure to danger during that phase as well. This makes it one of the safest ways to make isk in the game - but seriously time consuming. CCP might have done better to reduce the amount of time it takes to run repeated operations (as opposed to setting things up) thus allowing more time for the carebears to be at risk (like their other designs seem to want).
Oh well. No one ever said CCP's design team was consistent in sticking to principles.