Allow to modify nested parameters with LightningCLI

Nobody has claimed this yet.

### Bug description

Given the config file `my_config.yaml`

```yaml
model:
param:
sub_param: true
```
when I use lightning cli via `python train.py --config my_config.yaml --model.param.sub_param=False`

I get the error message:

`Unrecognized arguments: --model.param.sub_param=False`

### What version are you seeing the problem on?

v2.0

### How to reproduce the bug

Create a simple train file `train.py` that uses Pytorch lightning cli.

### Error messages and logs

```
# Error messages and logs here please
```
`Unrecognized arguments: --model.param.sub_param=False`

### Environment

Current environment

```
#- Lightning Component (e.g. Trainer, LightningModule, LightningApp, LightningWork, LightningFlow):
#- PyTorch Lightning Version (e.g., 1.5.0):
#- Lightning App Version (e.g., 0.5.2):
#- PyTorch Version (e.g., 2.0):
#- Python version (e.g., 3.9):
#- OS (e.g., Linux):
#- CUDA/cuDNN version:
#- GPU models and configuration:
#- How you installed Lightning(`conda`, `pip`, source):
#- Running environment of LightningApp (e.g. local, cloud):
```

### More info

_No response_

cc @borda @carmocca @mauvilsa

Reproduce the issue with the referenced train.py, my_config.yaml, and LightningCLI invocation. Start at LightningCLI's argument parsing and configuration override handling, then verify that the nested model.param.sub_param override is accepted and covered by a regression test.

Written by the indexing model from the issue text.