Minerals from Ward's are of two types, hand samples and research mineral samples. They are chosen for their relatively large size and euhedral morphology. There is no information about their chemical composition and really nothing can be assumed about their homogeneity.
An ideal microanalysis standard is free, has large grains, is homogeneous on both a micron and macro scale, and has been characterized by a method in addition to EPMA.
Most materials being considered for EPMA standards fail one or more of these criteria, usually homogeneity is the problem.
The Smithsonian Microbeam Standards are a far superior set of materials because they have been hand picked to identify clean grains, mounted and sigma ratio measured to assess intragrain and intergrain homogeneity, and analyzed by wet chemistry. Some of them are only available as small grains.
The wollastonite that John discusses is probably the best case material. It contains trace/minor Fe, Mn, etc. which is typical of many wollastonites (the crystal structure does not welcome more than trace concentrations of Mg, Fe, etc. compared to the augite and orthopyroxene structures). It may be a decent standard but it really requires a major effort to characterize the material that would be selected for distribution as a standard.
I would expect that the hornblende, like Kakanui hornblende, contains inclusions. That is the reality of natural minerals. That presents a real problem in stating what the composition "is" and issues regarding use of that material as a potential standard.
I support exploration of these materials. The ideal set of calibration materials for the geology community is end-member, stoichiometric, homogeneous materials available in large chunks. For example, I use Elba hematite as our primary Fe standard but always include secondary Fe standards in the PFE run. This Elba hematite is probably from the Caltech mineral collection, is not an internationally recognized EPMA standard, but is Fe-rich and homogeneous. The Smithsonian magnetite is an equivalent standard and could be used (the Smithsonian ilmenite is good but contains some inclusions of hematite and other phases). The Rockport fayalite standard likely contains grunerite and magnetite in the mineral separate used for the SMS wet chemical analysis, so it is decent material but the accepted analysis is faulty; this can only be corrected by cleaning the separate and reanalyzing the bulk material.
So the ideal standard setup is to have materials that have the highest element concentration expected in a probe run, but avoiding problem materials (Quartz: beam sensitive, Si peak shift vs. most minerals(?), Corundum: conductivity issues, peak shift, and so on). The accuracy of the calibration is then demonstrated by analysis of accepted standards such as the SMS materials (we use Kakanui hornblende).
The Ward's wollastonite could be a usable standard but I recall seeing BSE zoning in the large pieces. This ultimately, like many materials, disqualifies it because when you crush it for distribution, you have many grains with variation in the chemistry. When a user mounts up 1-2 grains they implicitly assume those grains are the true composition when they probably aren't. This issue is a problem for the commercial standard mounts but notice that they mostly contain end member phases like Sb2S3 which are stoichiometric and homogeneous.
The SMS are free to anyone who requests them. There is material from a larger size fraction that can be requested but that is not necessarily what was used for the wet chemistry, and especially for Kakanui hornblende, the larger fraction has grains with larger inclusion of ilmenite, etc. So there is no easy answer (except synthetic glasses, which I can elaborate on).
Cheers,
Paul