campaigning and debating and advertising were, for a while, designed largely if not exclusively to influence voting preferences (well, and to raise money of course). turnout was relegated to a quite ground operation in the last week or so of the campaign. it was known to be something any decent campaign had to do, but it was never seen as a deciding factor.
in that environment, polling with a question that amounts to, "if voting were as easy as just answering the question right here, right now, who would you vote for?" made sense.
as polarization took root, political operations found out that fewer and fewer minds were actually being changed. enough were to keep the polls moving about, but eventually they realized that turnout was starting to make more of a difference, and was easier and cheaper to affect, than actually changing voters' minds (and hoping that they turned out).
so campaigns and debates and advertising became focused on turnout from early on -- negative, slogan-based, appealing to fear, loyalty, and identity and based on often trivial or fabricated red-meat-for-the-base "issues" rather than anything that resembled actual debate on actual issues.
in this newer environment, turn out matters more, and asking who you would vote for doesn't shed much light on that key question.
however, you can ask if they've already registered, if they've missed and elections in the past, and when it gets close to election day you can start to guess the weather (which is well known to affect turnout).
and of course we know the republicans are well aware of many tricks to further discourage or prevent people from voting, such as removing names from voter rolls, voter id (poll taxes), voter challenges, voting machine fraud, shorter voting hours, fewer voting locations, fewer voting machines, etc.
= new reply since forum marked as read
