As for the potential scale of the problem, as of September 2010, there were 13.9 million trips in the month for Calgary. Not considering seasonality, this is approximately 170 million trips annually. If, for example the misallocation in resources was $0.50 trip (just pulling a number out of the air), then this would be a misallocation of $85 million ANNUALLY. From a very high level, I do think this is an issue that requires better data collection based on the potential size of the problem.
I would argue actually that fare differntiation actually provides more freedom of choice as opposed to social engineering. A flat fare, doesn't allow people who might use transit otherwise in reverse commute routes to do so that may otherwise use up that capactiy.
As for the problem of political palatability, the following is a good quote:
"Here lies a challenge to the public relations officers, who have the potential of earning a lot of goodwill from price differentiation. From other sectors, including rail and aviation, we know that customers appreciate increased freedom to choose between price and quality combinations."
As for the issue of "white flight", I would argue that this was a problem that was worst in cities like Detroit, but less of a problem in cities like San Francisco, Boston, and Washington, the latter of which all have differentiated pricing schema.
INVENTIVE PRICING OF URBAN PUBLIC TRANSPORT
Nils Fearnley - Institute of Transport Economics, Oslo, Norway
http://www.aetransport.org/lc_files/files/LPT-H-03%20Fearnley.pdf
On a personal note, I would class myself as a 'peak marginal' user. I have a variety of different options to use, some of which like bicycling are actually faster. Since, it costs much more to build the capacity for peak demand, every time I use the system I am actually needlessly burdening others who are travelling on backhaul routes and non-peak periods to supply that capacity.
The following is a simple and intuitive example illustrates this further: Assume an operator with almost zero marginal costs and two passengers whose willingness to pay in one direction are $1 and in the opposite direction are $3, respectively. If the operator were to charge one uniform fares then $3 would maximise profits,
but only one passenger would be served. With price differentiation, the operator could charge them up to $1 and $3, respectively, which would yield a higher profit
and provide affordable services for all customers.
So one can see this isn't also about improving profits per se, it's also about improving the welfare of people, and increasing the utilization rate of our existing public transit infrastructure.
I think political acceptability is dependent on getting the right information first, doing good research and focus groups on implementation, and then proposing to the public what the problem is (quantifying size of problem is important) and how it can be fixed. Do a horrible job at public relations, and the idea would not be politically viable despite it's potential for economic and social welfare benefits.