The inlib/histos are not AIDA compliants and inlib/exlib can't be seen as an AIDA implementation. The relationship to AIDA is only through the .aida XML file format that we can write and read for the histograms (by using inlib/histo/waxml and exlib/raxml with expat). At the level of the code, the OpenScientist osc_batch and osc_vis kits are still our AIDA implementations. And the relationship of OpenScientist with inlib/exlib is that future releases of osc_batch/osc_vis will intensively use inlib/exlib as a handy "core foundation" layer.
More explanations : the inlib/histos classes do not inherit AIDA::IHistogram and have no direct connection to any facility (storage, graphics). This is on purpose. We want to have classes that handle "core histogramming only" and arrange that connection to facilities or other products be done through other classes. The inlib/histo/waxml and exlib/raxml are examples of that (along with exlib/2root to relate to CERN-ROOT). The osc_batch is another example of using the "naked core histograms" in the context of providing an AIDA implementation. In this context the BatchLab/Histograms, that inherit AIDA::IHistograms, uses the inlib/histos for the core logic of histogramming, but have also code of its own to handle the management of histos in an AIDA::ITree (that can provide storage facility by using the AIDA API logic).