I don't think that second f is justified from what you've said. The "only if" seems likely in the modern environment but not guaranteed.
Even the "if" half doesn't quite seem guaranteed, though. If it's always easy to know which choice maximizes utility, then having more options is clearly better. But I don't think that's the case often enough to be confident in it. If someone is skilled at pretending to be "good", but finds themselves with choice paralysis in situations where genuinely "good" agents would proceed with dispatch at only a slight penalty to the average quality of the ultimate choice, you might be less successful for having the "dark traits" despite covering them up. And even that is ignoring the cost of doing the covering.
I don't think that second f is justified from what you've said. The "only if" seems likely in the modern environment but not guaranteed.
Even the "if" half doesn't quite seem guaranteed, though. If it's always easy to know which choice maximizes utility, then having more options is clearly better. But I don't think that's the case often enough to be confident in it. If someone is skilled at pretending to be "good", but finds themselves with choice paralysis in situations where genuinely "good" agents would proceed with dispatch at only a slight penalty to the average quality of the ultimate choice, you might be less successful for having the "dark traits" despite covering them up. And even that is ignoring the cost of doing the covering.