It seems to be possible to complete the survey without being a registered user of MN -- or at least, if it is impossible there is nothing onscreen that suggests your results haven't been submitted.
Given the number of cyber-politicos who want to use talk forums as a way of influencing opinion during the election, does this not mean that the survey is open to abuse? The kind of abuse that effected (e.g.) the Today Programme's Person of the Year poll, etc). They won't be here this time, perhaps. But once there has been a headline saying '40% of MNers will vote Labour da da da', they will be all over the next one.
And if MN restricts the survey to registered users only, the same politicos will register with MN in order to vote. This is one of the ways that I'm concerned that excessive pursuit of publicity under the heading of the 'Mumsnet Election' will damage the site. Usership statistics will become less and less a guide to the size or nature of the community -- and the poll results themselves will be meaningless as a source of knowledge about MNers. Subsequent political livechats will be dominated by people who are only on MN to sway the perception of their party.
And a proportion of the people who join in order to poll will stay in order to canvass. Even though it is hard to draw a line betwen canvassers who are here as community members and those who are here purely instumentally, I think there is a legitimate distinction to be made, and I worry about canvassing being accidentally encouraged.
I'm not against publicity-seeking as such (Thready here again). But I do think there should be lines drawn about how actively, and in what ways, and with what results for the site, it is encouraged.
Plus, even without being swamped by politicos, the survey will not be a scientific reflection of anyone's opinion. It will be another resource for journos to add a little bit of 'colour' to their election coverage: it often seems to be the case that the Mumsnet element in journalists' election coverage is part of the colour and packaging of news, not part of the political discourse itself.
As such, the survey is likely to contribute to the white noise that detracts from constructive political discussion during the election campaign. Good article on that here.