How to make life easy for an NKG developer to write NGINX configuration

[pleshakov]

Background:
At this point, NKG generates a non-trivial NGINX configuration including using NJS to delegate advanced request routing decisions. The config generation code is here https://github.com/nginxinc/nginx-kubernetes-gateway/tree/e06546993afb57fa59ae9b26427d190f65f3c3d3/internal/mode/static/nginx/config

Problems:

• Onboarding new developers - how to quickly learn how NKG generates NGINX config and why it does it that certain way.
• Helping experience developers on the project too (ex. I myself asked this question a few times - how many servers do we create a TLS listener and why we use those cryptic server_names? (server_name ~^; and a few times forgot the answer).
• Understanding guidelines and recommendations for extending NGINX config to add a next NKG feature. Ex. as a developer, how do I ensure the config I add is reliable (prevents NGINX errors), secure (doesn't introduce vulnerabilities), performant (doesn't slow down the data plane), consistent (what are the rules for variable names, formating, line length, etc.), doesn't break existing configuration?

Related -- we have a related #562 , but the point of that issue to cover information from the user perspective.

How can we address those problems?