`reportInvalidTypeVarUse` superfluous when using default values?

Code sample in [pyright playground](https://pyright-play.net/?code=CYUwZgBGD20NoA0IF4IEsB2AXAugCgEMAnAcwC4IAbNAZy0RxQgDpWBKCAWgD4rb6EOCq2YAoIA)

```python
def foo[X = int](arg: list[X] = ...) -> list[X]: ...
```

`pyright` complains that

```
warning: Type variable "X" may go unsolved if caller supplies no argument for parameter "arg"
  Provide an overload that specifies the return type when the argument is not supplied (reportInvalidTypeVarUse)
```

But that seems superfluous, given the default value. `reveal_type(foo())` produces `list[int]` as expected and not `list[Unknown]`.

I mean, I can see that maybe one still wants to get a warning here, but it feels more like a lint than an actual typing error.

I would like to silence such errors globally, but `reportInvalidTypeVarUse` also covers other "real" errors like incorrect type var scoping.

Start with the linked Pyright playground example and compare the reportInvalidTypeVarUse diagnostic with reveal_type(foo()). Determine whether a default argument should suppress or separate this diagnostic from other type-variable scoping errors; done means the behavior and configuration match the issue's intended distinction.

Written by the indexing model from the issue text.