feat: add ability to not "clean" target directory during retrieve

[techfg]

Currently, when executing a retrieve, the target directory is cleaned entirely prior to the writing the retrieved files. This has several shortcomings especially when a filter (e.g., app name) is applied.

I think there are a few options:

1. Make the default behavior to not do a clean unless a flag is passed (e.g., --clean)
2. Leave the default behavior and add a flag to not clean (e.g., --no-clean)
3. Make the default behavior to do a clean when no filter is specified and not do a clean when a filter is applied providing a flag to override the default behavior as appropriate (e.g., --mode [clean | no-clean]`

Generally speaking, performing destructive operations by default is not a good thing. That said, if you retrieve, then delete something on the server via web ui, then deploy (without doing another retrieve), the entity you deleted will be created again.

I think it really comes down to what perspective the CLI is intended to operate in - is it to "make the state of my local match the server" or is it to "update/merge my local state with the state of the server". The answer to this would likely come from usage statistics, user persona's etc. in terms of what the "default" would be based on the most common usage scenarios.

Regardless of answers to what the default behavior is, the user should have the ability to either clean or not clean to provide the appropriate amount of flexibility for all scenarios.