Is/ought revisited

Here is a reputedly defensible version of the famous principle that you can't derive an 'ought' from an 'is', taken from a paper in defense of the principle that 'ought' implies 'can' (Peter B. M. Vranas, 2005) (this was linked by ephphatha in a previous post), although the link is now broken:

(I/O) No valid argument has a conclusion that is a singular* moral claim and premises that form a consistent set of nonmoral claims.

My question is, does the following example serve as a counter-example to I/O? (Also, if so, why? If not, why not?):

1) Jones believes that Smith ought to concede that we can derive an 'ought' from an 'is'.
2) Everything that Jones believes is true.
3) Smith ought to concede that we can derive an 'ought' from an 'is'. (from 1 and 2)

*singular moral claims are understood as moral claims expressed by asserting or denying that a specific agent S has (or does not have) a moral obligation to @.