Should historians be able to make arguments to the public like "Roman society declined because luxury diminished the old republican virtues" (just an example), or is that no longer appropriate because it is inherently unscientific?

The question is whether or not the evidence seems to support that argument. That's an argument about change over time and as a historian you are constantly making just such arguments, but you must always be based in the evidence.

Now, one's personal bias is likely to play into your analysis of the evidence to some degree. In fact, I think the "modern discipline", such as it is, acknowledges that bias has realized that the quest to eliminate bias is unattainable. It is worth minimizing bias to the degree to which it is possible in order to approach the evidence as best you can, but you can't do it perfectly. Therefore, it is essential to make your biases known explicitly to your reader or audience.

That being said, I think I still probably haven't offered a satisfying answer to the real underlying question you're asking here. You might be asking something closer to "should the Historian make an argument that it was a bad thing that roman society declined, that it declined because luxury/the loss of republican virtues, and that the Romans should have tried to avoid that?"

To that, I have two further answers:

First, insofar as the historian is doing history there is little (perhaps nothing) that is explanatory by making such an argument. It just doesn't make your historical argument about what happened and why it happened stronger to stake out such a position as part of your historical argument. In fact, it is a rather ahistorical argument. I think this realization has made historians apprehensive about expressing these kinds of things as core to their historical work. The fact of the matter is that the history is better without it.

Second, I have absolutely no problem with historians, as individuals or as professionals, making such arguments in public as long as they are well reasoned and argued. In fact, I think historians do regularly make such arguments. I certainly couldn't teach the Soviet Union without engaging people with what the Soviet Union can teach us about authoritarianism generally speaking.

I think historians and people in the humanities in general are still working out how best to balance these two things. Pretty much everyone in the humanities that I've interacted with - many hundreds of professionals - would recoil at the idea that their work is entirely amoral. We do tend to believe the work we're doing matters and that often times it matters for moral reasons. But we also realize they the best analysis of our topics also tends not to be based in those moral judgements.

There is one more major wrinkle here.

Insofar as historians' professional advancement tends to come mainly from their historical work/publication, there is relatively little professional motivation to engage in this kind of public discussion regularly or on a wide scale. This is an institutional barrier rather than theoretical or methodological one that I think also plays a rather large role. If a professional historian wants to advance in their department and field, they are better off spending their time publishing more journal articles and attending more conferences rather than publishing more think pieces on morality in the New Yorker.