passing non error HRESULT to winrt::throw_hresult results in an assert, should this be made easier to diagnose?

[ChrisGuzak]

I was testing error diagnoseability and mis-wrote my test

        winrt::throw_hresult(ERROR_INVALID_STATE);

should have been

        winrt::throw_hresult(HRESULT_FROM_WIN32(ERROR_INVALID_STATE));

Rightly, this triggered an assert, but it took me a while to figure out I was to blame. A comment suggesting this possibility would have caught my attention.

Is this worth addressing?