I like this idea, I'm not too sure what the exact % of ridership is based on area, but if most ridership is crossing three borders than this can work. if the data does in fact show that a majority of trips are done locally then I think this may be a money-looser.
My rationale is (and admittedly I haven't run the numbers, but I was thinking of using an origins-destinations survey that hopefully breaks it down by travel mode) is that the combination of: increased fares for long distances (an extra $0.50 for long haul trips), increase in the cash fare to $3.50, and increased local travel within zones will make up for the loss of some of the short trips subsidizing the long ones. The 1 zone trips may be paying half of what they're paying now, but if double the number of short trips are made, it works out to the same. And of course you'd get the occasional person who forgets to tap off and ends up paying more than double what they should have paid.
I personally think the biggest jump in transit use that we would see would be off-peak short haul trips in the inner suburbs.
Id prefer either 2$ for your first entrance to any TTC vehicle, then 1$ for each additional transfer...
I would think that it would be easier to do a voluntary reduction in fare than ask people to pay an extra $1 after they've already paid their fare. Getting a reduction afterwards has the psychological effect of "alright, I got money back!". I don't think that asking people for money once they've already paid would go over very well.