Yeah, I can see that.
But the amount of still renders aren't the problem. Last time DPC was stating for weeks his update is ready and he's waiting for the finish of his animations in queue. Sure, no one has to believe his words. We can't prove them anyway. But personally I don't think he was lying there. So, if you want to make guesses you should do it on his animations, not his still renders.
Theoretically:
EP1. 1000 screens, 1 animation, takes 100 days (thus 1001)
EP2. 1100 screens, 2 animations, takes 110 days (thus 1102)
EP3. 1200 screens, 3 animations, takes 120 days (thus 1203)
etc.
Instead of 1001 it could be 10000001, or whatever weight whatever variable has.
The days are the denominator. The size and number of variables in it (screens, animations, QA, testing, free roams, etc.) just change the value, whose numerator
is relative to all other episodes, as they all simply increase. Additionally, the denominator which is known, also increases. Notice how the weight of the screens, or value of the animation #, or # of more variables are irrelevant. We don't care because we're not projecting how many free roams or this or that he'll have. Screens are simple and accurate enough.
The ratio of the numerator increasing will obviously provide a more accurate trend, with more variables, to explain what it is we're measuring...but, again, we don't care about that. We care about time. Not what was done or how or their weights. Even if you magically want to measure how DPC does 1 free roam, 1 mini game, 1 this, or that for technical reasons to be even more accurate...all that means is your window is going to be slightly smaller for your projection because you have a more accurate sigma numerator.
The end result is the pretty much the same, as you saw with my 2 attempts at trends, which both calculate to Method 1's 576 or Method 2's 577 days.
