It's meant as a rule of thumb. That level of precision is unnecessary.
Generally speaking, industrially-processed foods are going to be overall less healthy than less-processed alternatives. There are a multitude of reasons this may be so, and between them there's enormous variance in both the size of and our confidence in those effects, not to mention between specific products.
Being overly specific just invites pointless quibbling over those details. Of course, not being specific just started this pointless quibble, so damned if you do damned if you don't.
> Group 1 - Unprocessed or minimally processed foods (fruit, vegetables, eggs, meat, milk, etc.)
Group 2 - Foods processed in the kitchen with the aim of extending their shelf life. In practice, these are ingredients to be used in the kitchen such as fats, aromatic herbs, etc. to be kept in jars or in the refrigerator to be able to use them later.
Group 3 - Processed foods. These are the foods obtained by combining foods of groups 1 and 2 to obtain the many food products for domestic use (bread, jams, etc.) made up of a few ingredients
Group 4 - Ultra-processed foods. They are the ones that use many ingredients including food additives that improve palatability, processed raw materials (hydrogenated fats, modified starches, etc.) and ingredients that are rarely used in home cooking such as soy protein or mechanically separated meat. These foods are mainly of industrial origin and are characterized by a good pleasantness and the fact that they can be stored for a long time.
Nobody really knows for certain what things do to the body even in isolation. For one example, see Yellow #5 [1]. It's associated with all sorts of nasty stuff, but actually proving a causal relationship is not really possible. So some countries ban it, some require special scary labeling for products with it, and in other countries it's A-OK.
And processed foods aren't in isolation. You're mixing up all sorts of stuff which can viably result in chemical and other changes which will vary by product, producer, and even time. Trying to pin all of that down is not really possible.
I'm specifically not making that claim. (I also don't disagree with that claim) I'm making the claim that those activities are indicative of a constellation of activities that as a whole are categorically unhealthy.
Because doing that is an invite to the field of weak/conflicting evidence because food science just don't have this level of granularity yet. Though I believe to be undeniably that the replacement of ultra-processed cheap foods with real food is beneficial. I'm too lazy to research it for the sake of an argument but I can support it at least on a personal/anecdotal level. It's not scientific but as a rule of thumb if you don't know how it was made better to not eat it.
