callModule handling is broken

[dgkf]

After a long hiatus, trying to rerun the callModule handling functions has now been spawning errors. Given the spurious nature of the bug fixes over the past year, I think I have inadvertently broken some of the recursive module server function handling.

Targeting a timely fix to set the table for module design discussions.

[dgkf]

After taking a look at it, this stems from my misunderstanding of the session$makeScope() functionality in shiny, which I had been misusing to try to get access to a scoped nested module server function body's environment.

Will have to do a bit of exploration as to how to gain access to a nested function environment to try to determine if a callModule result extends the getInitializationCode generic.

[dgkf]

I hesitate to call this 'fixed', since it will ignore getInitializationCode for anything but top level module calls.