make it clearer how and when to set engine resources

[ryandawsonuk]

As a follow-up from #753

If the request size is large (e.g. binary images) then the engine pod likely needs more resource dedicated to it. By default the engine will currently have access to the node's resources but this will not apply on clusters where a default is set.

We should have an entry in the docs explaining this and how to control the resources