Lambda is not a ratio. It is a number that tells you how far from Stoichiometric the tune is. It does not care what fuel you are running and all AFRs use Lambda as the root and are then multiplied by some number specific to that fuel.
If you start mixing fuels (i.e. methanol injection) then you can throw AFR out the window as it's useless.
If you learn what Lambda is and run or tune to multiple fuels, you will never tune to AFR again. Trust me.
I can understand the value of lambda to a point but would you really consider it as the only viable solution for a multi-fuel tuning scenario? It seems that this assumes that for given load point, a common lambda value would be targeted for gas, methanol, or ethanol... or any combination thereof. Meth injection on a gas engine will definitely "throw off" an AFR expressed as a gasoline AFR, but I've not found that it does so to the point of worthlessness.
There are a few ways that folks describe or visualize Lambda. One is that it is the stoich ratio for a given fuel normalized to 1. Given this, I am not sure how observing measured AFRs as Lambda reduces or eliminates error when mixing fuels. The O2 sensor measures dissolved oxygen content in the exhaust stream and has no idea what fuel was burned to create the exhaust. Unless a wideband controller knows the ratio of the two different fuel masses being delivered to the engine and somehow accounts for this in its measurement, I don't understand how Lambda is perceived as a solution. What would be measured and expressed as a Lambda of 1 (or any other value) will change as different ratios of fuel combinations are applied to the engine.
In short, in a changing world of fuels, each with different stoich ratios, Lambda is NOT a constant. As long as you are comfortable tuning to the scale in which your AFRs are being displayed, that's the key.