I think we need to clearly distinguish two things here, though. A lot of software engineering ritual is dedicated to shipping software - that is the part that tends to be of very remote relevance in research and scientific contexts. But there are also more fundamental principles and techniques that concern writing code, which apply regardless of the reason why you're doing that, and are beneficial in any context - in a sense that, if you follow them, you will spend less time overall writing code (and thus more time doing whatever it is you're trying to achieve with that code).

However, the problem is that those things are not trivial and have to be learned before they can be applied. And that is where the value becomes questionable - the time spent learning all that could instead be spent on learning other things. But if you already have a software engineering background, you do benefit from doing things "right" to some extent.