Yes, that is a pitty. These attributes were used for workarounds.

Second it, these attributes used for demographic information for closed access surveys. That is key to analysis.

Either it is anonymous or it is not.
There can't be a mode that allows to undemine anonymity by entering name and email in the attributes.
You can't promise technical anonymity if there is no.

The real pity was, that this - using attributes in an anonimous survey - was allowed in some releases.

I second this statement from joffm "You can't promise technical anonymity if there is no."
The pity is that every surveysoftware began to offer such an mode. But LimeSurvey is so depend on workarounds that changing setting will break workflows.
So I want a token remove mode. I want that the token is not saved in the results but i want to use personal attributes for workarounds.

The market research survey know-how will be gone soon. Employee surveys conducted with anonymous mode but results are not aggregated enough to protect people.
Rules of thumb applied without looking at crosstabs which reveal peoples identity in small groups.

Exactly. Anonymous mode should remain anonymous. There is no disagreement. Nobody wants to pretend something is anonymous when it isn't.

The feature, as jelo suggests, could be an addition to the standard survey mode, letting the admin select in "Participant settings" which participant fields are usable in the survey and will be captured in the results table.

Something similar already exists at the tail end: The "Export results" feature offers the user "Participant control" to decide which participant fields are included the exported file - just that the data to be excluded shouldn't even have been captured.