The IETF process is open and there should be not private input to it.
That being said, as editors, WG chairs, ADs, or the IESG we occasionally get input that is directed only to us, not to the relevant working group or the IETF community. As an example, as an editor you may discuss with someone interested in the topic and get feedback. Or the IESG may get last call feedback that is sent only to the IESG because the comments are either too trivial (editorials, indications of support) or because they want to highlight a issue that the sender does not want to discuss in public.
How should we take these into account?
The first response is that we should in almost all cases direct the person to provide their feedback to the working group.
There are only a couple of possible reasons why this might not be the case.
Sometimes we get trivial feedback, like those editorial comments that we can just take into account.
Sometimes we get background information (such as indications of support), which absent any major controversy do not have to be public.
Sometimes we get information that for privacy reasons (e.g., disagreement with employer) may not be publicly aired.
However, every time there is a technical, substantial issue or an issue where the WG is divided, decisions cannot be based on private input. Comments need to be out in the open (directly or indirectly) and the community needs to be able to discuss them openly. Community consensus rules, not chair or AD opinion.
Reasonable suggested changes to a document do not necessarily need to have been discussed in public before they are made. But the changes do need to be properly exposed and accepted by the WG i.e., on the WG list. In doing so, it would be normal to expose some form of motivation for the change. (If privacy is a concern, that motivation need not expose the individual concerned.)