Context
Aconex, Autodesk Construction Cloud, Bentley iTwin, ProjectWise, these are mature CDE products. They all support 'open standards' in some sense (IFC import/export, API access). The difference between them and Weaver is structural, not feature-level. The question is whether the data shape is public, not whether there is an API.
An API on top of a proprietary data model is not the same as an open data model.
Explanation
Three structural differences:
Internal data model
Proprietary CDEs store data in their own schema, with their own object types and their own relationship semantics. Export to IFC or CSV is a translation layer, lossy by definition. Weaver's internal data model is RDF, structured to ISO 15926-11. There is no translation layer because the on-disk format is the open standard.
Extensibility
On proprietary CDEs, customising the data model usually means asking the vendor for a roadmap item. On Weaver, it means editing RDF configuration. The metamodel is yours to extend, in the moment, without waiting for a release.
Replaceability
Proprietary CDEs are designed to make migration painful. Open-standards CDEs are designed to make migration boring. We treat data portability as a feature, not a contractual concession.
Proprietary CDEs are good products. For assets you'll operate for ten or fifteen years, they may be the right choice. For assets you'll operate for fifty, the open-standards architecture is the only architecture that survives the vendor risk.