Ultimately, the structure of the infobox is up to the language head. Which ones are included are up to the heads of each language, like wikipedia.
I think we should not add fields until we have data for them, for example, there isn't any data for "cpu use index".
I think "helppatch name" should only be used on the objects that don't use the standard naming scheme (i.e. *-help.pd).
Also, the "messages" section might be redundant. It might make more sense to including the messages in the "inlets" section, since occasionally, different inlets accept different messages. (Most of the time, messages with selectors are sent to the first inlet).
I also think that it should be the same structure in all languages. I don't see much point in each language going astray depending on the mood/resources of its participants - for that there's the puredatainfo already.
But while checking the infobox template, I saw that those extra fields were already programmed, that's why I asked if they should be added to the template. In that case, I would suggest to remove those fields from the infobox structure for now - new fields can be added at any time, after they've been "approved".