"Ultra-processed" may be a little fuzzy at the boundaries, but at least it's a specific enough term that we all know and mostly agree on what we're talking about when we use the term. UPF-ness is a heuristic that can help determine whether or not a certain food is junk, but once you've categorized something as "junk food" you've already decided it's unhealthy. No one has to study or debate whether or not junk food is unhealthy.
And the value judgement is the only part that has an especially meaningful connection to health. It is admittedly really difficult to do rigorous research on the health impacts of diet, but the connection between "ultra-processed-ness" and health outcomes is super messy. And even when we can get data that suggests negative health outcomes, what's the actual cause? Preservatives? Lack of fiber? Hyperpalatability? Nobody has been able to clearly articulate a cause and if we could then we could just focus on that thing rather than the "ultra-processed-ness."
It's not a question of whether or not they fit that definition, it's that the definition itself is so expansive that it allows equivocation between food products that are meaningfully different in their ingredients, health outcomes and environmental outcomes.