Against the view that reasons can be propositions and needn't be true propositions, it seems to me that such a view faces a difficulty similar to the difficulty faced by views that identify pieces of evidence with propositions and drops the truth requirement. Suppose there's a reason to drink the stuff in the glass and that reason is that there is gin in the glass. Suppose there's no reason (available) to not drink the gin in the glass and so we say:
(1) You ought to drink the contents of the glass.
If pressed for an explanation, wouldn't this be a true explanatory statement:
(2) You ought to drink the contents of the glass because there is gin in the glass.
But, doesn't (2) entail:
(3) You ought to drink the contents of the glass and there is gin in the glass.
But, (3) is false. There's no gin in the glass.
In general, it seems that one of the things that normative reasons can do is provide part of an explanation as to why certain things ought to be done or oughtn't be done. But, explanatory statements like (2) are factive. So, the only way I see to save the view that severs the connection between reasons and truth is to say that the question as to whether some normative reason is a true proposition or false one depends upon whether it enters into explanatory relations or not. Sort of a weird view, don't you think?
Myself, I don't have the semantic intuitions that let me say things like:
(4) Among the reasons there were to drink the contents of that glass was that there was gin in it, but of course there was no gin in it.
Some have them. Alright, but doesn't everyone pretty much take it for granted that 'p because q' is true only if p is true and q is true?