Which brings us back to the difficulties with different definitions of terms as used by different disciplines...
Something about "parsimony" way earlier in the thread. :-)
But seriously, the only thing I remember about Brownian motion was reading Einstein in translation, ages ago; and a writeup of a molecular dynamics study using "simulated" Brownian motion to attempt to incorporate its effects on solvent caging of a substrate at an enzyme's active site. So your "everywhere continuous" and "nowhere differentiable" fail to ring a bell...Although it does present room for speculation about the size or scale of the system (number of particles and detail of interaction potential) during which a bunch of discrete particles can begin to be successfully modeled as a continuum. Could you please post a reference to a link or two? Enlightenment gratefully accepted. :-)
My considered opinion is that Einstein and Feynmann are two of the most elegant and economical writers in English I have come across (except J.R.R. Tolkien).
All I know about Brownian motion is in Ito calculus:a primer here