Defining success as billions of users gets tossed around a lot, but why is that the dogma? If you can make a product smaller than that and have it be profitable, then what's so bad about that? Is it really so hard to manage that apart from Google's gigantic products?

If success is going to be billions of users then yeah, probably do option 1, because nearly nothing is going to get there. Why not have a pipeline to spin these products off or have them survive with lower expectations?