I work in the health sector and honestly, it comes down to a couple things:

1) They have more data. It's not clear that it's better.

2) That difference is, to the eyes of many of us, showing up, making things worse, and then "pivoting".

3) I get to decide what I want to do. I want to add a project on X? I go work on it.

I would seriously quibble with the anarchic independence you describe in #3 because (depending on the field certainly) you need to fund that work with grants and postdocs etc... Do you might feel like you can but it's generally largely dependent on convincing other groups to fund that work.

I'll Grant that The above doesn't really apply to pure math or philosophy

I'll admit that I managed to find a position with an abnormal degree of freedom in a fairly applied field (epidemiology) but I don't think that's any more atypical than people with particularly nice jobs in the private sector.

But right now, I'm working on projects in emerging infectious diseases, healthcare associated infections, some philosophy of science stuff, some algorithmic work in network science, and something that can best be described as "digital humanities".

I have a friend who’s a physicist-turned machine learning researcher working on health issues. I asked him about why he decided to say an academic researcher instead of going to the industry where (to me) there’s clearly more data. I was told that academic institutions tend to land better partnerships that companies don’t tend to get maybe due to privacy concerns.

I think your friend's impression is pretty accurate.

Industry in healthcare, especially the tech industry, has a long road of trust building ahead of it.

I think we're also a little jaded. We've been talking about how big, real-time data streams like pharmacy sales data, social media, etc. are going to change the game for disease forecasting since I was an undergrad. We're still not even close.

> companies can go from research to product that ostensibly makes a difference at scale with a speed that absolutely no University could.

Isn't that the point of being an academic? That you don't have much, if any, interest in generating a product?

I guess I didn't make my point clear.

For researchers at University of Toronto as well as within Google, neither (probably) wants to directly work on a product.

However both would certainly (At least in my experience) like their work to impact humanity on a wider scale then simply the number of citations their paper has. Again primarily thinking about applied researchers here.

So in that sense, both researchers are insulated from the producing of something business related based on their research. However one of them has a significantly greater chance of their research being used to actually affect people at large scale in their lifetime.