Except they will charge you the same extortionate egress fees for outbound traffic from a single compute instance ("an apple"), no global content delivery network involved ("a diamond").
It doesn’t matter where your instance is, the point is the traffic will flow on Google’s worldwide private backbone until it gets very close to where it’s going.
This is similar, with different tradeoffs, to what Riot did for their gaming backbone. They wrote a blog about this but I cant't find it.
Instead of relying on consumer ISP's to route gaming packets for speed, they would instead route them for costs. Taking extra hops and extra latency before reaching the target server.
Riot built an overlay network using their own fiber and negotiated BGP deals with consumer ISP's to get gaming traffic off the ISP's internal network on onto Riot's ASAP so they could route the packets directly to their gaming servers.
Thanks. I think those cover details of what I'm talking about but not the blog post I remember.
A key issue discussed was some setting relating to BGP (or similar) that consumer ISPs wouldn't initially implement or wouldnt configure for Riot. Something related to weights or something that affected the cost analysis for routing decisions that prevented packets from wanting to take the Riot path.
edit:
Just fully read that QZ article. I think it does a good job of describing what they did with useful inforgraphics (esp the "Actual Data Route" graphic) for a layman audience. I think many of us here can extrapolate what kind of effort they had to go through to get that done.
