The OCTO Technology blog is written by engineers at a French consultancy, covering software architecture, APIs, databases, and web technologies.
HN commenters cite it for its hexagonal architecture write-up and a Protocol Buffers benchmark, in discussions of layering and efficient data formats.
For further reading, Dan Abadi's post at http://dbmsmusings.blogspot.com/2018/09/newsql-database-systems-are-failing-to.html addresses the Spanner issues more directly.
I don't think there's a great reference for the Percolator transaction architecture specifically, but there is a brief description here: https://blog.octo.com/en/my-reading-of-percolator-architecture-a-google-search-engine-component/
One thing I learned recently is that FoundationDB is essentially the Percolator model (timestamp oracle), even though FoundationDB predated the Percolator paper.
I am currently a big fan of the hexagonal architecture which is pretty close to what you describe as far as layering.
https://blog.octo.com/en/hexagonal-architecture-three-principles-and-an-implementation-example/
https://blog.cleancoder.com/uncle-bob/2012/08/13/the-clean-architecture.html
I also advocate for packaging by feature rather than layer (each feature defines its layers)
https://hackernoon.com/package-by-features-not-layers-2d076df1964d
Then use compression. gzipped json is highly efficient over the wire.
https://blog.octo.com/en/protocol-buffers-benchmark-and-mobile/
The primary purpose of binary data formats is that they can be parsed quickly and some even allow direct access without a parsing step.