itsatdm, I can answer you from my experience, but then application into the auto/bike world will be extrapolation.
In our industrial natural gas engines the wide band sensor operates over a very wide range and give a good stable signal. The narrow band are inexpensive but have a very narrow range, in fact it is slightly off if you use natural gas as fuel.
For a catalyst where emissions are very important you have to have a very specific mix of O2, NOx, CO, HC to have the catalyst reduce all the pollutants. When our fuel mixture changes and we get more ethanes, butanes, and propanes mixed with the methane it changes when the mixture burns. Most of these extras make the mixture burn faster and the piston is closer to TDC when the peak firing pressure hits. The combustion is hotter and more NOx is produced and less CO and HC. The window that the catalyst works at is verry narrow, like .955 to .957 AFR (or some number like that) And this difference in the combustion products is enough to make a marginal catalyst quit working.
So, the first difference here, is we have removed the catalyst, so we don't have to be that precise with our AFR. Now we are wondering how far the ethanol changes the mixture and its rate of burn speed. In the end our ECU can compensate some for the fuel and makes sure we have the same amount of O2 in our exhaust, but it cannot compensate for the combustion speed. This might slightly throw our engine off of best power, or best economy. How much, I don't know. My guess would be that the ethanol will slow the burn speed. Timing could be used to compensate, also a leaner mixture would burn faster (provided you are not already at stoichiometric). In the end I don't think ethanol will throw you off too far, unless you are already at the ragged edge.