CKAN Groups - Editor or Member not able to add Group to Dataset (DDS-642)

Right now neither CKAN Group Editors or Members can add Groups to Datasets unless you are an editor dataset owner's org.

We need either the Editor or Member profile to work for MVP, with the acknowledgement that the last bullet under Editors would be an enhancement.

**Ideal states**:
* **Group Members abilities**
- Add Groups to Datasets
- all PUBLISHED or PENDING ARCHIVE datasets
- if an editor or admin of an organization, include DRAFT and PENDING PUBLISH datasets for those organization(s)

* **Group Editors abilities**
- Add Groups to Datasets
- all PUBLISHED or PENDING ARCHIVE datasets
- if an editor or admin of an organization, include DRAFT and PENDING PUBLISH datasets for those organization(s)
- _ENHANCEMENT_: Edit all Group properties (description, image and privacy checkbox) other than that the Group Title and Name (URL)

* **Group Admin abilities**
- Add Groups to Datasets
- all PUBLISHED or PENDING ARCHIVE datasets
- if an editor or admin of an organization, include DRAFT and PENDING PUBLISH datasets for those organization(s)
- Edit all Group properties (description, image and privacy checkbox) other than that the Group Title and Name (URL)
- _ENHANCEMENT_: Edit all Group properties (Group Title, Name (URL), Description, Image and Privacy checkbox), requires Keycloak Delegation Management

RISK
* Not applied in Beta: Those assigned Group "membership" in cat classic will not be able to test.
* Not applied in MVP: Those assigned Group "membership" in cat classic will not be able to add groups to datasets, therefore we will have to do it on their behalf.
* Granted Group Admin, means they would be able to change the group name, which then they would break their own access.

The issue names no files, tests, or entry points. Start by tracing the CKAN UI flow for group membership and adding groups to datasets, then confirm whether Editor or Member access is the MVP choice. Done means the selected role can add groups to the specified dataset states while preserving organization-based restrictions.

Written by the indexing model from the issue text.