Performance is the primary benefit for adopting the author's approach, actually.
Normally the bottleneck of scaling an application is data access and storage, there aren't typically a lot of problems that are CPU bound.
Adding additional layers doesn't increase performance, it increases latency and processing time. The closer you can keep your processing logic to the actual data, the better your performance is going to be.
That being said, there are reasons why you may not want to follow that architecture, but generally performance isn't going to be one of them.
Normally the bottleneck of scaling an application is data access and storage, there aren't typically a lot of problems that are CPU bound.
Adding additional layers doesn't increase performance, it increases latency and processing time. The closer you can keep your processing logic to the actual data, the better your performance is going to be.
That being said, there are reasons why you may not want to follow that architecture, but generally performance isn't going to be one of them.