I'm not quite following one part of your calculation. I think you are off by twice as much. Here's how I do it, and it has been pretty close in practice:

Say for 2000cc of fuel injector, you jet at 25%(500cc of meth) and spray 100% meth. I don't use ideal stoich AFR, but power AFR, which is 12.5 for gas, and 5.5 for meth. So meth is about 44% of gas, for simplicity I tend to round down to 40%.

So if you jet at 25% and use 100% meth, that's like 10% more gas(40%). 10% more gas is roughly 1.0AFR in most situations.

Now for jetting at 20% and using 50/50 that's 10% methanol, which is about 4% more gas, or .3 to .4 of an AFR change.

On paper this may look funny, but we have been pretty accurate at predicting AFR change and providing proper jetting over the years.

NOTE: One thing to remember is the margin of error on whatever dyno you may be using. Some will have a issue with a .2 to .3 repeatability so many runs and proper load bearing dynos are needed to measure accurately.