There is an upstream and downstream component to this issue. I will tackle upstream first.
On reading the docs added by mthoemme, these look good, but I think we can make it more user friendly by splitting it out into modules / subsections.

e.g.

• Autoscaling
  • Concepts
    • KPA
    • HPA
    • Editing the configmap
      • CLI (downstream only - this is the only option for upstream)
      • console (downstream only)
  • Configuration options KPA
    • Scale to zero
    • Termination period
    • Concurrency
    • Scale bounds
  • Configuration options HPA
    • Scale to zero
    • Termination period
    • Concurrency
    • Scale bounds

Open questions

• Is HPA == BYO autoscaler, or does this need to be another section under concepts?

TO DO:

• Tackle TBC docs: https://docs.google.com/document/d/1ubszZxikwe0O1JsA-Rjwdv3YA-rgHWKLI9WPAD083ig/edit#heading=h.6gkfbgz3jup9
  • related github issue https://github.com/knative/docs/issues/2375