I think the problem is that two different notions of expert have been widely confounded within the present scientific training structure, let alone within publi

I think the problem is that two different notions of expert have been widely confounded within the present scientific training structure, let alone within public perception. There is a huge amount of prior literature on many topics, so to be an expert in something like immunology you're almost definitely a PhD, or maybe a very academically inclined MD. However this notion of expert simply means you have learned intricate details about many of the important previous results, most of which are probably true. It does not mean you actually have the ability to do good science in the field of immunology. But for any technical question in your niche that isn't uncharted territory you can serve as an expert. Such expertise has obvious uses in many other domains, and as there isn't really a great term for it, it gets called scientific expertise. Perhaps the person may be called a biologist, but that also has scientific undertones which may or may not be warranted. I'm sure you are aware of this, but I don't often see it explicitly stated, and so it feels like people talk past each other in a sense. It is wholly possible to approach e.g. chemistry in a non-scientific way, just as something like business can be approached in a scientific way. I don't think the double language will be resolved any time soon, since I don't see much effort to make this distinction in the actual design of PhD training - forget about random internet forums.