You could probably do it by using two channels and programming the slave channel to follow a non-linear response. I have done this to get flap travel to match very closely.

Or you could not worry about it. I have seen posts by Ed Moorman showing exactly how much difference a model will tolerate. It appears to me that the engines will self-synchronize in the air as the slow one windmills up a little and the strong one is dragged down. Currently flying a model with twin OS .52 4-strokes and one is almost 1000 rpm's stronger--other than a little pulling to one side in a loop it doesn't seem to effect flight.

Love the Hornet!! I've got to build another some day.
Chuck