I think that any fuel tweaks you make other than warm up ones will just adapt out (fairly quickly).
cf's without wmi are ok for the occasional >6 when the cars hot, and 3-6s when cool, but if you run wmi, if you limit it to 3's, when the wmi fails (which it will) the knock sensors have enough headroom left to save you from disaster.
Water slows down the flame front speed, so you can ignite the charge earlier without the max pressure rise occurring too early (before the piston is travelling down), and get more of the pressure rise over before the exhaust valves open. It also increases charge density (mass) so there is more mass expanding and pushing against the piston.
Maximum power is achieved if when you take the absolute maximum expansion (and consequently temperature) out of the charge. Low egts=max efficiency from charge. When you ramp up charge density, and rpm, the time available for the charge to burn/expand is reduced, egt's climb, efficiency drops. The sooner you can ignite the charge without knock, the more power can be extracted on the expansion stroke.
Cooling the intake charge has the same effect (intercooling), so wmi can be win win.
The third benefit is that the meth is additional fuel (v high octane) which needs twice as much air to burn, and messes with lambda measurement a little, but at 5% addition will not do any harm, and gives a good octane improvement.
Fueling is generally over rich to provide charge cooling where inter cooling is insufficient. Max power will be made somewhere in the 12.5-13:1 region, but most over fuel to provide charge cooling when turbos and their inter coolers are pushed too hard. That's the beauty of wmi, no need for over rich (power sapping) af ratios.