I use turn out in the loosest sense of not voting.
It's true that turnout among people with postal votes is high - 77% in 2011 - but in past elections we don't know to what extent postal voting represents something of a commitment to vote rather than a more convenient way of voting.
I would have thought that for someone not committed to voting, actually returning a postal vote requires about the same level of motivation to have someone from RIC register for you and then to see the postal ballot form, open it, find a pen, mark the cross and then walk to the post box with the form in its envelope. RIC did half the work - they took the mountain - but having reached the trough, the horse now has to drink. To wreck two perfectly good metaphors. A fair proportion of the newly registered still won't vote.
In 2011 about 6% of postal ballots were rejected mainly because people mucked it up - not returning the ballot paper or the statement with their signature, or just not signing it. I'd expect more of the non-voters who vote to muck it up and have their papers rejected.
The other factor that will bring turnout below the 80% reported in surveys is that at least some of those people still aren't registered. They might think they're certain to vote but they're not.
I'm sticking with my estimate of 73% turnout.
I'm not aware of anyone asking how people intend to vote. I expect most to start asking whether people have already voted and how.