The only conclusion I can draw from this is don't bother buying a high aperture lens (i.e. fast <f2) for a DSLR. You are just throwing money into a hole.
The data presented does not support that conclusion.
The text accompanying the data
suggests that as a possible future finding, but as yet it is unsupported by fact.
With the best modern cameras, a 1/2 stop increase in ISO (what the graphs suggested) has near-zero effect on quality. And the
suggestion that depth-of-field is adversely affected to a large degree, is just that: an unproven
suggestion.
Definitely something worth
keeping an eye on, but so far it's just a sensationalist article.
Cheers