I had a quick look at your extensive nanorisk framework document, and have a few comments. I apologize when some of them might already be addressed in the document since I did not have time to read it in full detail. So take my comments with this caveat:
What seems to be missing with respect to the issue of hazard evaluation is a discussion about the usefulness or – on the opposite end – nonsense of using in vitro studies. The focus seems to be solely on bioassays, which is fine, however, one could save unnecessary expenses and time by doing simple in vitro – without and with cells – assays. These may be good for identifying the real "bad actors". At least I think you should say something about not using in vitro assays if that is what the group decided. Again, though, chemical reactivity should be part of the characterization of a new nanoparticle in my view.
Also, the concept of having well defined – physico-chemically as well as toxicologically – reference or benchmark materials should be included, both highly reactive as well as benign.
RE: the bioassays, there seems to be no discussion on the issue to using realistic dose/concentration ranges which is an extremely important issue and ought to be addressed – at least its principle.
With respect to agglomeration/aggregation, measurements are done in what medium? Water, saline, plasma, all would be important, in particular also when it comes to determining as to whether a nanoparticle can be "de-agglomerated". Although it is already a lengthy document, more details would be helpful.
With respect to risk management: Did you consider the use of warning labels? Certainly, for most nanoparticles there will not be a significant risk, but for some it might. Not knowing for a new product as to whether it is benign or of significant toxicity, how should it be treated? People working with it should be informed somehow.
These are some brief comments, again I apologize that I did not have time to read very carefully the whole document – so these are just some initial thoughts.