While interesting, it would be nice to see actual code. Also, without more information on the specific code for both sides, and the testbed in place it's hard to really judge.
In the end, the numbers between the two weren't significant enough to inform a clear decision for most use cases. Also, the whole point of microservices is to be able to scale out workloads. Use what's best for a given task/microservice. Duh.
Agreed. The bias is towards Go from the start. Additionally, the example isn't entirely trivial; it's dangerous to leave out half of the argument. The Node code could suffer from a lack of optimisation, notwithstanding the fact that the engine version isn't disclosed — and that makes a huge difference.
The notion that Node is aimed at MVPs is a hackneyed one, too. Node powers plenty of massive production sites & apps, and is omnipresent in cloud infrastructure (Lambda, Lambda@EDGE, Google Cloud Functions, Firebase etc). The point has been raised and mooted incessantly, and quoting Ryan Dahl — who resigned from the project during its infancy in 2012 — is uncontextual and misguided in equal measure.
Posts like this one hearken back to Steffen's question here: https://www.echojs.com/comment/25768/1
As they're essentially X vs JS opinion pieces on a forum specifically intended for JS.