It might but perhaps not that much.
Firstly, whenever a library can provide the same functionality as a language feature, the former tends to be favored. D's `unittest` -- while clean and convenient are far necessary.
Regarding macros, C++ is gradually seeking out superior solutions in cases where macros are used currently. A good example here is finding an alternative to macros that use `__LINE__` and `__FILE__`[1]. Generally, I would expect many of the disadvantages of C++ test frameworks to be addressed over time with general solutions that extend far beyond this one use case.
Some people are very used to separating tests from the code that is being tested -- often into an entirely different project root directory. This can be an obstacle to acceptance of a feature which aims to improve locality -- as D tests tend to.
Then there are the alternatives. `static_assert` does not require additional frameworks to function and does not even incur a runtime cost. For the diminishing number of tests that cannot be performed at compile time, `assert` and contracts[2] -- while not solely intended for the purpose of unit testing -- play a significant role in testing code and ensuring correctness.
So overall, I would not be too hopeful for such a proposal per se but I'm in favor of improvements that head in the same general direction.
Thanks,
John
1. wg21.link/N4519
2. wg21.link/P0542R1