old polling methodologies out of the window?

Just a brief post this time. Polling companies must be sweating - results are suddenly much more volatile and from chugging along nicely a week ago when all the pollsters had a hung parliament with the Tories 10-40 seats short, we're now all over the place with predictions.

I'd be interested to know which pollsters are asking respondents:

• if they watched last Thursday's debate
• if they are registered to vote (the last date to register has now passed)
• if undecided, a few more probing questions about style vs substance

Either way, the volatility means that there may only be some polling winners, with some big losers. Will we have a better idea of which companies are the most reliable until after the event?