If a research does not release the code used in their research, their papers should not be trusted. We wouldn't trust a paper where the researchers hid their experimental methods or analysis - it wouldn't even be allowed to be published. But if their experimental methods or analysis are done in code, those parts are allowed to be a black box and we're supposed to trust them. If we were willing to just trust them, there wouldn't even be a peer review process and it certainly wouldn't deserve the imprimatur of 'science'. This has been a problem for years, and I will personally not be the least bit surprised if we eventually see a death toll attached to it.
