I didn't ignore it, I addressed it implicitly with the reasons I rejected the validity of NP to software developer comparison; but to address it explicitly:
The median RN plaisibly makes in the same rough range as the median junior dev (“junior dev” being a seniority level within an occupational rather than a distinct tracked occupation in the BLS data, this is hard to tell with certainty, but plausible.)
OTOH, “junior dev” is the entry level of software development, beyond which people are expected to progress with a few years experience. RN is the full working level in nursing; the median across registered nurses includes a much wider experience band than that across junior devs, so they aren't really comparable: at the experience level of a median RN, a junior dev wouldn't be a junior dev any more.
> I'm also trying to figure out what your point in the third paragraph is, exactly.
You claimed that looking at Chicago rather than national numbers would shown that the latter suffer gross distortion in tech wages due to tech hubs like SF. In fact, the relation between the nursing sub-group you chose to focus on and software devs in Chicago is very similar to the national relationship (both are a little under the national level in Chicago, and software developers a few percentage points moreso than NPs, but not enough to make a significant difference in the relationship.)
> Again, that's the whole point - if you look only at national medians or averages, you're completely ignoring the effects of regional outliers
But I'm not looking at only the national numbers, I just compared them to the numbers for your chosen non-tech-hub locale and found that the gross distortion you claimed in the national numbers, compared to non-tech-hub locale, doesn't exist.
> If you can't do that, look at non-tech cities and countries for an indicator of what unskewed salary comparisons look like between professions.
But I just did that with you selecting both the non-tech-hub city and the non-tech profession, and the national vs. local comparison wasn't materially different. (I suspect that the reason is that locality has only modest effect on the median and a lot bigger effect on the total number of positions and the shape of the high tail; if we were looking at means rather than medians there'd probably be more distortion.)
However, I have pretty big issues with BLS data in general, so it's pretty clear we're not going to come to an agreement here based on that data set. If you're going to treat BLS data as gospel and other sources as garbage then we can save time and stop here.