By the Numbers

[StellaTerra]

Hello,
Over the last two months (leading up to my first WC build) I've done a lot of research. The most striking and interesting website by far is Martin's Liquid Lab. Unfortunately, the website is layed out in such a way that much of the information is inaccessible to neophytes like me, but I think I've managed to piece together the larger picture of how the loop works (in terms of the physics) and what all of those nifty looking graphs mean.
The thing that struck me just yesterday is that, if you knew enough about physics, you theoretically should be able to make a sort of free body diagram (not sure if that's the right term) for the thermal situation. What I thought of was this: What if there was a way to create a mathematical formula for the performance of each of the pieces of hardware, and then piece them all together such that you could look at a series of independent variables and tweak them to get an accurate estimate of the end performance? Then, if you had that, you could make a website where people go and punch in a list of information (what radiator, what pump, how much tubing, of what size, etc.) and know what your processors temp would be. I told a friend of mine about it and he described it as the Theory of Everything of Water Cooling.
This is what I know:
The temperature of a component = ambient temperature + delta water temp + delta component temp
Theoretically, given the right set of data, the deltas should be consistent across all similar model devices. We would need to know:
ambient temp
water temp
delta water temp
component temp
delta component temp
system TDP
component TDP
head
static head
friction head
flow rate
fan CFM

And then the three most difficult factors to determine:
radiator performance behavior equation
pump performance behavior equation
component water block performance behavior equation

Some things wouldn't be accounted for, like heat dissipation inside the case, but those could either be compensated for, or assumed out of the equation (given that this is only an estimate anyway.)
Martin's flow-rate calculator gives me an idea that something like this can be done, seeing as there already exists something that can do a large chunk of the work.

So, yeah, this is probably crazy and impractical, but I wanted to see if anyone else thought this was possible.