I don't see the 'direct progress' you mention, then again maybe my sarcasm detector is miss-calibrated.
They transitioned from lemons to limes without realizing that they had gone from a kind of citrus that protected from scurvy to one that didn't. Their interest in meat was due to expeditions which unambiguously demonstrated that lime juice did not protect against scurvy while polar bear meat did. And no simple categorization would have explained to them why you don't want to use copper in preparing food.
Witness for example Scott's ability to demonstrate that seal meat prevented scurvy, and his inability to realize that trying to preserve it by boiling it wouldn't work.
1 - Something about citrus fruits cures scurvy. Correct.
2 - Something about living on a diet of preserved foods causes scurvy. This is even more correct, since now we're getting at the cause.
But I agree we're seeing some interesting cognitive biases here.
Once their original theory is disproved, they focus on explaining the latest results, and seem to forget that they still have to explain why citrus fruit ever seemed to cure scurvy. (Or perhaps the story is more complex than we suppose, and they had some other explanation).
What's certain is that they don't test their ideas with enough experimentation. There seems to be a bias towards formulating theories, and trusting in the expert's judgment, without any empirical verification.