Use import dagster as dg in our examples and documentation

[PedramNavid]

In this PR, we've started using dagster as dg as an import: #23864 (comment)

The main benefits of this approach vs the traditional method importing objects directly into the namespace are:

0. There is prior art in many libraries of this approach, notably pandas, sklearn, numpy, matplotlib, pytest and many other packages that export many objects.

1. Fewer import lines in code samples/snippets. Compare the first image which is 12 lines shorter. This makes code easier for users to understand parse.

2. Another side benefit is that type-ahead experience is improved.

In VS Code, for example, typing dg.sensor and pressing Ctrl+Space provides you with all the exported Sensor-like objects

Compare to typing sensor instead with import dagster

3. We can always revert this change in the future by searching the examples for dg. if we decide this approach is fundamentally broken.

4. Easier to write docs/examples as you do not need to constantly add/remove imports from your code blocks as you develop.

We have also seen some slight positive community feedback.

Some risks include:

• Will take time to update all our examples, some examples/docs/blogs may contain the old way of doings things. However, all code examples will still work regardless of which method we use.
• Users who do not catch our new import structure may find they get import errors

There may be other benefits and risks I have missed. Opening this up for discussion before we fully commit to making this change.

[PedramNavid]

Closing the loop here -- given there has not been any push back, and have had generally positive reaction to this change, we will be making this change in our new docs site.