It's not like once we learn something in the lab we can just make it available to everyone for free. I mean, you could go to Wikipedia right now and read about, say, aeronautics and all the theory of how air flowing over a wing can provide lift, and articles on structural engineering and so on. But having that knowledge is not the same thing as having a plane, or having access to a plane ticket. If you would like to fly from A to B, you need a functional aircraft first.
I am obviously not arguing against research, and I think your whole deal of arguing that there's some kind of tradeoff to be made between them is wrong. You can have more research and more economic growth. In fact, if you don't have more economic growth, it'll be harder to have more research, and to some extent vice versa.
This isn't to get embroiled in a real tangent about network neutrality. Nobody is advocating for any kind of particular regulatory framework for the internet, here.
I am (un?)fortunate to have my first memories around the time of the first private networks being connected to the internet (late 80s).
Back then, the internet was run by academics.
It wasn't until the 2nd or 3rd round of commercialization efforts that it really became compelling.
So the structure of the internet is because of the academic beginnings, but the actual content (data, apps, etc.) are IMO b/c of desire to acquire wealth.
The Internet as you know it is also a direct product of its academic background. That's why we have had network neutrality so inherent in its culture. That's why there are only peers, not producers and consumers. If it had been designed as a product, the network itself would never have been designed to be dumb with smart endpoints. (And there are countless examples of this.)
If Tim Berners-Lee designed the web to be a product, there would be a marketplace of competing web-like products, which in turn would have given us other features, but it would never have become societal infrastructure in the way it did.
Sometimes we need academia and science, sometimes we need products.
The question here is how to optimize between the two.