Is that engineer reporting the $750,000 salary statistically significant? I feel that based on the averages excluding this salary it would be best to view that data point as an outlier.
A single data point is never "statistically significant" on its own. What you probably mean to ask is whether its standard score is likely assuming a standard normal distribution of values [0]. Without running the numbers, I would guess not.
I know someone in HFT whose base salary is $150,000 but got a 10x bonus (he made $1.65 million) last year. Average bonus in his HFT are 2-10x base salary he says.
any decent stats analysis has built in outlier detection and ignores them, there are people that either lie or just make stuff up for attention and detecting these (data trolls) is just part of the analysis. Blindly averaging is never effective.
Why would they all be data trolls who lie or hope for attention?
Fwiw, 250k+ seems fairly standard in some fields I've interacted with. You don't need to look very far to find them either. A good Oracle DBA will make that or more in a reasonably large corporate environment; so will a good SEO hacker in the gambling industry insofar as I've interacted with them.
It does beg one question though: how high in the corporate hierarchy were the positions with, say, $360k/year or more? If the salary is that of an IT exec in large corporations, it's not necessarily comparable to the salary of IT staff underneath them or of the consultants they might hire. (Only one thing seems reasonably sure: they weren't working for early stage start-ups.)
Why? It's not impossible to think that there is an engineer somewhere out there contributing seven figures a year to a company's bottom line . . . why shouldn't they capture a significant part of that productivity?
[0] https://en.wikipedia.org/wiki/Standard_score