There were a question, a reply, and a comment to it ...

... which turned out to be a little too long for a comment, so here it goes.

The construction in question (There is an N and an N) is, I believe, one of those places where the grammar appears to be slightly embarrassed because it doesn't quite know which way to go; or rather, the grammar isn't quite sure why it has done what it just did.

One succinct way of putting it appears in Huddleston 1984:69 (Introduction to the grammar of English):

'... the normal rules of subject-verb agreement do not apply to the there construction: person-number inflection in the verb is shared between there (which is 3rd person singular, like it) and the NP following the VP.'

That is, the verb wants to agree with singular there, while at the same time it also wants to agree with the following NP. In the present case, the following NP is of the 'an N and an N' type, where the presence of 'an N' immediately following the verb swiftly pulls it in the singular direction, deciding the fate of this grammatical tug-of-war. By the time the plurality of the whole NP is drawn to the speaker's (sub?)consciousness, it is too late.

It is a bit tiresome to have this kind of number concord game between there and the following NP each time you say 'There ...', so the growing grammatical tendency is for there to win. It is often said that there's (but not quite there is) is becoming like a grammatical marker to introduce an NP, singular or plural (as in 'There's a lot of nasty people out there, you know.').

Jespersen 1924:155 (The philosophy of grammar) gives a wider view on the matter:

'Whether or not a word like there is used to introduce them [i.e. existential sentences], the verb precedes the subject, and the latter is hardly treated grammatically like a real subject.'

(He then cites Danish and Italian (as well as English) examples to illustrate the point, before further expanding on the issue; quite illuminating. Jespersen is joy to read.)

To sum up, in a there-construction, the verb looks in both ways. When the following NP is clearly and powerfully plural, it wins the number concord game. When it is clearly singular, well, there is no need for a number concord game.

When the following NP is 'an N and an N', there wins, with the help of 'an N' - but the speaker, realising the plurality of 'an N and an N', is sometimes left with a funny feeling: um, what did I just say?

And I love that funny feeling.