Competent systems engineers are in charge of high level decisions in theory.
In practice, they might be less in charge than they should be (so that, more or less criminally, managers prioritize business objectives over safety and quality) or less able to do their job and less competent than they should be (probably because better processes and higher standards would mean spending money to cause "problems").
It's never any engineer's job to make decisions/define constraints. That's on the business side whether it be management, marketing, analytics teams, etc... I seriously doubt any engineer decided this. The decision to go with a single non-redundant system and use software to improve safety margins seems like a cost cutting measure which definitely seems more like something that would have been mandated as one of the design goals. Maybe engineering presented various options or something like that, but I find it very hard to believe the risk decision laid anywhere else other than very senior management.
