If you're really motivated to participate in x you should try, if you're a bootstrapping solopreneur you may have no choice, but if you're looking at mitigating hiring risk, a month trying to code/sell/design outside your comfort zone doesn't really move the needle on figuring out how to evaluate programmers or whether your startup actually needs a sales team. (If anything, you're more likely to get the wrong answers from the DIY approach because maybe the sales function actually would be beneficial if it was carried out by someone who could sell...).

That's where you get more insight in less time from leaning on people in your network who actually are marketers or salespeople or product people to get a view, especially if they're not financially incentivised to propose a particular route.

If you want to mitigate the risk that ICs need structure and resources, you focus your hiring on candidates that have worked without structure and resources. But even having a codebase written by someone learning not to rely on Scrum meetings and JIRA tickets to prioritise is still better than having one written by someone learning to code.