Suppose Bob does the objectively right thing -- what he has most reason to do -- but he does it for the wrong reason. (Say he defends an innocent person from a violent attacker, but only because it gives him an excuse to be violent himself.) Bob is clearly blameworthy here: he acted maliciously, after all. But we might ask more specifically whether he is blameworthy for the action, or just for the malicious desire that motivated the action. Doug Portmore, in Commonsense Consequentialism (and at PEA Soup) opts for the latter. Even more strongly, he claims (2.11) S is blameworthy for freely performing X only if S does not have sufficient reason to perform X, all things considered. (Note that there's no requirement here that the agent be aware of the reasons in question. Their mere existence suffices to excuse his action.)
I have two objections to this view:
(1) If the agent is merely blameworthy for having the malicious desire, then it shouldn't matter (so far as their blameworthiness is concerned) whether they act on it or not. But this clearly does matter a great deal. Having malicious desires might make you a bad person, but questions of blameworthiness concern a subtly different kind of evaluation, more concerned with what you've done. Bob is blameworthy for acting on his malicious desire, not just for having it.
(2) Whether an agent's X-ing is blameworthy or not is surely an intrinsic matter, i.e. depending only on the internal states (including volitions) of the agent, not on purely external matters. Two intrinsic duplicates, who make all the same decisions (and enact all the same volitions) must thereby be ['originally'] blameworthy for all the same volitional exercises (actions).
[N.B. This is compatible with moral luck affecting what external consequences their blameworthy action renders them 'derivatively' liable for, and hence the degree to which they are blameworthy -- see my tripartite analysis of responsibility.]
Portmore seems committed to denying this platitude, since external differences could make it so that there was sufficient objective reason for Dupe1 to X but not for Dupe2 to do so. In that case, (2.11) implies that even if both agents were enacting exactly similar murderous volitions, only Dupe2 would be blameworthy for exercising his agency in this way, for Dupe1's action unwittingly aligned with "sufficient reasons" -- roughly, what God would advise, given the broader situation.
Why does Portmore make this claim? He points out that (i) the capacity to respond appropriately to reasons is a precondition for moral responsibility, and (ii) it makes no sense to blame someone for flawlessly executing a capacity that's a precondition for moral responsibility. So an agent can't be blameworthy when they flawlessly execute their capacity to respond appropriately to reasons. But I'm not sure how Portmore gets from here to the conclusion that an agent can't be blameworthy when they perform an act for which there happens to be sufficient reasons. After all, an agent might perform the advisable act from sheer luck, rather than as a result of competently exercising their rational capacities. If they act from malice, and it's just luck that their action conforms to the demands of reason, why couldn't it still make perfect sense to blame them for their flawed exercise of agency? Things may have turned out well, but damn, they did a poor job as an agent. (One may exercise incompetence in a job or role, even if this fortuitously causes the 'goals' of the role to be better achieved.)
P.S. I should note that while Portmore uses the premise 2.11 in a larger argument for moral rationalism -- the view that it can't be morally wrong to do what you have sufficient reason to do -- he might do just as well with an appropriately modified version of the premise, guaranteeing blamelessness for X-ing to agents who not only have sufficient normative reason to X, objectively speaking, but furthermore act from those very reasons, i.e. as their motivating reason. Anti-rationalists seem likely to fall afoul even of this weakened principle.