If we can persuade the banks to provide pages which, when visited generate random but guaranteed fake account info which can be entered into the phisher's website, then those of us who are alert enough to recognise a phishing email can do something proactive whenever we receive one.

1) This gives us a warm fuzzy.

2) The more dud data there is in the phisher's database, the more attempts he has to make in order to carry out a successful transaction.

3) As soon as the bank sees an attempted transaction using the known fake data, it can identify and block the phisher's IP address, rendering even the real data in his database useless.

At the moment I'm reduced to making up phoney account data in the name of Scrooge McDuck, which is hardly ideal, only addressing points 1) and 2).