On the stack trace, I think this is what the modern "observability" stuff is all about; traces, wide events, etc. One event per request. DataDog will say they do this, Honeycomb will say they do it and that DataDog is kinda lying, now there's OpenTelemetry, it's a deep rabbit hole.

It's easy to say this is a lot of work to reinvent something you get for free with a (single language) monolith, but at least it's recognized as a problem worth solving.