Francois -
I certainly don't think a new WG is needed, rather we should start to look at how to deal with 3D data within DM and DAL. Currently this is more of a discussion topic than anything else. While we don't know yet what is required to address the general problem of 3D data, it is clear that it is closely linked to image and spectral data access and we should take this as a starting point.
The Characterization model is supposed to apply to any VO dataset which should include 3D data. All it does though is physically characterize the dataset in terms of bandpass, sampling, resolution, etc. - there are many issues of access to 3D data which this model alone will not address. But we would like to use the standard characterization model for all data including 3D data.
The Euro3D effort is very interesting, I was not aware of this. This looks like it could help greatly to define from the data analysis side what is required to deal with 3D data. My main concern is that it appears to deal mainly with O/IR spectroscopic instruments (IFU, MOS, Fabry-Perot, etc.) but we need to consider spectral data cubes as well, e.g., from radio or from multiband surveys.