ipympl does not work in JupyterLite

## Describe the issue

Due to a dependency on `tornado` https://github.com/matplotlib/matplotlib/blob/c06e8709dde6504d396349c0c80ef019c88c3927/lib/matplotlib/backends/backend_webagg_core.py#L22, ipympl does not work in [JupyterLite](https://github.com/jtpio/jupyterlite)

We might be able to implement a custom Timer https://github.com/matplotlib/matplotlib/blob/c06e8709dde6504d396349c0c80ef019c88c3927/lib/matplotlib/backends/backend_webagg_core.py#L118 based entirely on `asyncio` (which works on JupyterLite).

## Workaround

As a workaround, one could do
```python
import sys
import types

sys.modules['tornado'] = types.ModuleType('tornadofake')
```

Disclaimer, I would not be held responsible for anything that would go wrong if you copy this code!

Start with matplotlib's lib/matplotlib/backends/backend_webagg_core.py, especially the tornado dependency near line 22 and the Timer around line 118. Check how ipympl reaches this backend in JupyterLite, then verify that the integration works there without tornado and that interactive timers still function using the available asyncio support.

Written by the indexing model from the issue text.