My proposal was that propositions are a fusion of properties of a particular type. I'll illustrate with an example:
'John loves Jane' expresses the fusion of the following properties:
Being John&Loving, Loving, being Jane and being loved
I argued that this is a characterization that most accurately describes the meaning of the sentence, and thus is the best candidate for being the proposition.
This too is ambiguous. Consider the following:
The property of being Jane & being loved by John
The property of being John & Loving Jane
These are (prima facie) distinct properties. Each are equally good candidates for being the proposition expressed by 'John loves Jane'. Apply Bennaceraf.
One way to respond is by saying that english doesn't parse things this fine, and that most sentences are ambiguous between a few very related propositions. Another way to respond is to say that despite appearances, the two properties described above are the same. This is totally implausible, since one property is had by John and the other is had by Jane.
I think the moral of the story still stands. A benacceraf dilema only shows that there's a bit more work to do describing the proposition (fully). It doesn't show that there is no proposition. Moreover, we might not even need a full description of the proposition, a partial one will often do. If that's the case, proposition theory stands, and the benacceraf dilema doesn't support any non-existence claims.