From Google's point of view, it's worth it. If they think they could develop this in-house for $5 million, and maybe it'd be great, maybe it'd be okay, and mayb

From Google's point of view, it's worth it. If they think they could develop this in-house for $5 million, and maybe it'd be great, maybe it'd be okay, and maybe it would suck. Spending $50 million to let them have the pick of proven technology that they can already see working before spending the money removes the risk.