We use scalar at Microsoft. Working with massive scale of some repos would be near impossible with "vanilla" Git.