They didn't need to embrace standards or build APIs, they just had to hire people who knew how to build products that real people outside the valley wanted. Google got lucky with its first product (search), acquired a company to make it a successful business (Adwords), and kept buying companies to try and onboard innovation.
The Product managers are what count in building a new product. Hackers are rarely good product managers.
Academics are only a problem when they don't get user feed back before hand. A good researcher can build product that fits the needs of the users, and when paired with good UI people you get a winning product. Google isn't good at getting user feed back they work like Mathematicians, not like Anthropologists and Psychologists.
Math and Anthropology are still academics. But not in the same field.
The general impression I get is that they facilitate hundreds and hundreds of projects, let them develop, and if it looks good, let them loose into the wild and see what sticks. Most people only get to see the successful ones. I think the vast majority of these projects never see the wild.
In contrast, I get the impression that other big tech companies lean towards thinking something up, then trying to push on to users.
Perhaps I've just fallen for PR, but thats the impression I get.