Any task requiring or being > 100% better when done with more than one person is now a team. We are now optimizing team output and talking about low-skill is no longer an accurate description of the problem.
It also under attributes how important having a quality fitment/parts prep person could be to the quality and pace of the output. Adding another person can make the work 2x or more productive. If they can spot problems before they get to the welding stage, even more.
When I was young and while I was probably the A-team of the B players in the org, I romanticized about what it would be like to be in a high performing org of all A-players. This thinking is flawed, it devalues the importance of the b-players and puts an outsized importance on the high value heroes.
That prep person may be fungible on day zero, but by the end of the week, they should be a high functioning team member. The value bigot will say that the person that only preps the pipeline should only be paid X/hr because the work is low-skill. But their value to the system is much higher. We should be thinking in systems ant labels about what each component of that system is worth.
Nearly any able-bodied worker can do those tasks, and they don't require much skill. That doesn't mean they have no value to the overall process; obviously they do. But there's a big difference between finding and hiring a trained welder (or training someone up), and finding any random worker who can lift things and follow some basic instructions.
Absolutely agree with you that focusing only on the "value add" tasks can cause blindness. But that doesn't make the other "non value add" tasks somehow high-skill, high-quality, difficult-to-replace tasks.