Improve notebook cell model lifecycle

[msujew]

What it does

After https://github.com/eclipse-theia/theia/issues/13488, I've noticed a few issues in incorrect caching of notebook cells. Mostly related to them staying in storage for too long, thereby overriding new cells. This is mostly related to:

• Missing calls to $acceptEditorsAndDocumentsDelta when creating new notebook cell editors
• Dangling monaco-model references after calling createTextModelsForNotebook
• Dangling monaco-model references after disposing a cell model

How to test

I've been pretty successful in reproducing the behavior on master by:

1. Opening an existing notebook file
2. Adding a few new code cells to the start of the file
3. Closing the notebook file again
4. Opening it again

On master the new code cells contain content of other, existing code cells (this seems to happen almost at random, though it is consistent). This change should produce empty cells after opening the notebook again.

Furthermore, reviewers should test that cell execution on newly created cells works as expected (to test the newly added call to $acceptEditorsAndDocumentsDelta).

Review checklist

• [x] As an author, I have thoroughly tested my changes and carefully followed the review guidelines

Reminder for reviewers

• As a reviewer, I agree to behave in accordance with the review guidelines

[msujew]

@jonah-iden good catch! I actually already broke this in https://github.com/eclipse-theia/theia/pull/13606, but it should be fixed now.