Other LMS's also offer a partial credit option of equal weighting (or some other equivalent name) where the student would get an equal amount (say 1 out of 4 points) for correctly selecting an item that is correct (or True) and for not selecting an item that is incorrect (or false). Under that partial credit scheme, the answers given in the original post above would have 2 correct and 2 incorrect answers. He would have gotten 1 selected correctly and one unselected correctly resulting in a score of 2/4. This drove me nuts when I used Canvas at UCLA Extension and I worked around this weird scoring quirk by using Matching with 2 choices (Correct/Incorrect) to get the proper scoring. I used D2L Bright Space at CSULB for many years. It offered the equally weighted partial credit in addition to no partial credit and R-W that we have in Canvas for it's equivalent of the multiple answers question type. My only complaint was that equally weighted partial credit was not the default ...
Sadly, I have a bunch of image processing questions that result in binary 2D arrays as answers that I was able to use Multiple Answers to check where selected meant 1 and unselected meant 0. That obviously just flat out doesn't work using the Canvas R-W partial credit. Matching would work, but is definitely more cumbersome.
Given the lengthy time for which Canvas has not joined the other products in offering an equally likely partial credit option, perhaps the solution is to propose a new question type to offer similar functionality to multiple answers, but with equally weighted partial credit. Maybe "Multiple Selections," Multiple Checkboxes," or "Multiple T/F."