`--allow-redefinition-new` does not mean "new" implementation but "new" variables

[toslunar]

mypy/docs/source/command_line.rst

Lines 636 to 639 in df05f05

	.. option:: --allow-redefinition
	This is an older variant of
	:option:`--allow-redefinition-new <mypy --allow-redefinition-new>`.

mypy/docs/source/command_line.rst

Lines 644 to 646 in df05f05

	We have no plans to remove this flag, but we expect that
	:option:`--allow-redefinition-new <mypy --allow-redefinition-new>`
	will replace this flag for new use cases eventually.

The flag --allow-redefinition-new added by #18727 enhances inference of new variables, as the PR description says

Infer union types for simple variables from multiple assignments, if the variable isn't annotated.

This feature is rather orthogonal to the --allow-redefinition flag. In fact,

> mypy --version
mypy 1.18.2 (compiled: yes)

> mypy redef.py --allow-redefinition
Success: no issues found in 1 source file

> mypy redef.py --allow-redefinition-new --local-partial-types
redef.py:3: error: List comprehension has incompatible type List[list[str]]; expected List[str]  [misc]
Found 1 error in 1 file (checked 1 source file)

with the example in the doc of --allow-redefinition

def process(items: list[str]) -> None:
    # 'items' has type list[str]
    items = [item.split() for item in items]
    # 'items' now has type list[list[str]]

[ilevkivskyi]

Hm @JukkaL should we imply --allow-redefinition semantics w.r.t. re-assigning a different type in --allow-redefinition-new?