MY USUAL PLAYTHROUGH
For me in Vanilla, the game starts as follows:
1. Lay down starter base, with 1-2 labs and half yellow belt of copper / iron plates. For that, I usually have 3 copper 10 iron, and 1 steel smelters. If I get backed up, not an issue. I also allocate room for up to 160 boilers.
2. Design / blueprint the main base consisting of the following smelting lines, spaced out for electric mining drills. Stone x1, Copper x6, Iron x6, and Steel x4. Steal x4 lines are 24 iron and followed by 24 steel smelters. The stone, copper and iron are 32. I this time I allocate space for the raw resource bus from where I will set up the raw resource incoming train station to where the smelters.
3. Set up the main bus, which consists of 6 copper plate, 6 iron plate, 1 iron plate, 4 steel plate, 1 stone, 1 brick, 1 coal belt. There is lots of room on the main bus for other intermediary.
4. Plan for science to eventually be 2 science per minute, setting up red, green, and gray science (yes I prefer to use the color, rather then the name).
5. Set up the the first 3 utility lines producing belts, inserters, pipes, trains, train signals, rails, fluid tanks, steel chests, med and large electric poles and lots of other stuff.
Those first steps can be easily accomplished in a few hours, and then the slow down happens, since now I have lots of other things that need to get set up while building blue science, such as:
- Expanding smelting
- Building fluid rail yard
- Setting up refinery, including allocating space for fracking
- Getting the first train to run, most likely to an oil patch
- Depending on the size of starting resource patches setting up the first raw resource outpost
- Setting up base perimeter, with a real wall, turrets and radars
- Flushing out the raw resource train station.
- Breaking down the starter base
At this state of the game, oil is just a small piece refining of oil is just a small piece of the puzzle. It is the one which appears to have the least return on investment (ROI), and in my opinion the one which is the scapegoat. For the most part, refining is not broken, what is happening at this stage of the game, is the player's attention is fragmented on many other tasks, some of which are implied.
Once the this is set up, the game speeds up again, with purple and military science (if it was not yet done). During that stage of the game, there are less side tasks, which consists of mostly monitor what was built, pollution, raw resources, and expand as needed. Implementing purple science in .16 was a joke, and rightly re-balanced in .17. Even in .17, purple science, after implementing blue science appears way simpler and easier. It is also at this time that intermediary train station is built and the first logistics robots Kitty Hawk moment and take flight.
Gold science slows down, and it is mostly due the speed of construction of rocket control and the need of processing units for other items such as nuclear rockets, armor, and so on. Bottom line, the first rocket launch by this time is a non-issue.
After the first rocket launch, the next major slow down, is due to the lack of speed and productivity modules, but by this time the base is in transition to a mega base able to sustain 1 rocket launch per minute and beyond. What usually breaks a map is when with the player computer's UPS becoming an issue, sometime around the 25+ rocket launch per minute or 10+ science per minute range when using mods. Without mods, this happens much earlier.
WHAT I IDENTIFY AS THE ISSUE WITH PROPOSED SOLUTION
My key point is refining is a small piece, and there is lots of other items impacting the game, especially between the implementation of green science until that first blue science is put on a belt. When I look at the number of specified and implied tasks needed for blue science automation, I see a massive spike of sub-tasks
What I see is a possible solution is to split blue science packs into fluid science and processing science, call it, I'd call it cyan and dark blue. I don't see that as an issue, since we will have red + green, then cyan + blue, then purple + gold, then military + white. This gives us 8 science packs, which can be fit on 4 belts deliverable to to labs. Consider cyan being composed of plastic, solid fuel, and concrete, all requiring fluids for construction. Blue then becomes advanced circuits, engines, and power switches. Designed right, it can split what is required of the current blue science to provide a more visible ROI to the player. But a change of this scope best left for .18 or a mod.
The best interim change which breaks or disrupts the least amount is better tutorial for oil processing, and a warning for when storage for a fluid is backed up. As an alternate solution, we can provide a flare stack, an entity which burns off either heavy, light, or petrol fairly quickly and only generates pollution ... but why?
MY CONCERNS AND RECOMMENDATIONS
My concern with the proposed recipe changes is the amount of mods, and bases which will be broken with the proposed changes, especially with the recipe changes. The tech tree changes will be somewhat disruptive within the mid game, and in the long run it will not be mod or map breaking issue during the mid and late game in the same way that proposed changing the refining recipes will be.
I am strongly against implementing the recipe changes to any version tagged as .17. I like what V453000 did publishing the mod
https://mods.factorio.com/mod/Oil-Changes to let the players try out the changes and provide feed back. I am neutral on this change being implemented in .18 and beyond, mostly due to it being too early to start discussing this, in so much that we still do not have a first stable version of .17. as a new baseline.
Hiladdar