If making the most power possible is the goal, then yea, you're exactly right.
The problem with your direction overall is the risk. I don't know about the Focus, but on our Mustangs, the factory ECU doesn't have any way to know if the meth is spraying or not. If you tune it to take advantage of the meth (i.e. leaning it out and/or advancing the timing), and then you have a meth system failure, your computer doesn't know that the meth isn't spraying any more. And therefore, you wind up with a whole lot of timing, nowhere near enough fuel, and nowhere near enough octane. Boom. It's a HUGE gamble.
Tuning it via the IATs negates that risk. You only increase the timing if the IATs dip below what you know they'd be without meth. And the only way they dip below those known temps is if the water/meth is on. Therefore, the ECU has valid feedback on whether or not the meth system is working, and as long as it's tuned properly, it will be completely safe, even in the event of a failure.
Doing it this way obviously requires a decent sized dip in IATs when the meth is spraying; if the dip isn't big enough, it's hard to distinguish whether the dip is from the water/meth, or if it's from something else like cooler atmospheric temps. This is why I normally recommend more water than meth (~75 / 25), as the water has a much higher heating capacity than meth, and thus will absorb more heat, and result in a bigger change in temps. I ran 23-24 degrees of timing at 9.7:1 compression with a non-intercooled Eaton at 12-13 pounds, with absolutely no detonation, with washer fluid, which is about 75 / 25 water/meth.
Likewise, this is why I say it may not be worth it on an intercooled centri or turbo setup where the IATs are already decently in check. The water/meth may not cause enough of a dip to safely and confidently add any timing into the tune.
You're definitely right about the advantages of water/meth at the line. With my old SVO setup, I had a manual override button to spray the meth; I'd spray it a good bit during the burnout, and then I'd give it some short pulses once I was on the line and up on the converter. It did wonders for cooling the IATs off the line.
The Focus, like the Mustang, has no way of knowing if the meth is spraying or not. I used a failsafe device that required a WOT signal and meth solenoid signal. It would allow the meth spray without being WOT (since meth sprayed based on boost pressure) but not the other way around. If WOT was detected but no meth signal detected, it would kill the ignition to save the engine. I never had an issue with my AIS pump so I never had to use it. Nice thing to have though.
I tested a lot of different mixtures in various positions. I didn't see much difference in IATs between straight water and straight meth, (with appropriate and respectively sized nozzles) but I also didn't have to cool very much with the mixture either since I had the FMIC. I gained more and more power the more meth I sprayed though. While it is true that water draws out more heat, it also displaces the air in the charge piping and the cylinder. It absorbs the heat and holds onto it.
Water (H2O) has a very strong bond. While it only takes 212*F to turn it into a gas (steam), it takes extremely high heat (over 5,400*F) before it's molecular structure begins to break apart. Considering even nitromethane only reaches 4,350*F adiabatic flame temperature, you're not going to be able to break the water's bond in the cylinder. It passes right out the exhaust as steam.
Methanol (CH3OH) has an auto-ignition temperature of just under 900*F. The molecular structure breaks apart, oxygen and carbon are released into the chamber and accelerate the flame front. The heat that was captured in the charge piping gets released in the chamber. Another advantage is that methanol is considerably lighter than water by more than 20%, and also evaporates at a much lower temperature of just 148*F. If the air in the charge piping is above 148*F, the methanol boils and becomes a gas suspended in the air. This is why methanol works so well when sprayed pre-compressor because it will evaporate in the compressor housing, becoming fully suspended in the air, preventing heat from building up to begin with. Water, on the hand, will pass through the compressor housing as a liquid, clinging to parts, and more likely to cause erosion.
What I found to work best was 80% meth / 20% water by volume. This seemed to create a pretty good mix of heat reduction and octane boost.
In a car that's already running compression through the floor such as an SRT-4, more water will probably work better as the methanol will be of much less use. Same goes for a non-intercooled car that needs to bring temperatures way down. The higher the compression and combustion pressure, the more the methanol will shine.