From my time in a corporate research lab: The formal review process was conducted in parallel to initial review, with the goal to amend to prevent internal details from being widespread and amending for camera-ready. While people would freely review each other's papers, the goal was to increase initial acceptance rate. However, our lab notably, and perhaps exceptionally, mirrored the structure of an academic lab with corporate funding, even though all researchers were employed by the corporation.

Google's story would not fly where I worked, or where I work now.