There's been some discussion of the article on comp.lang.ada
My favorite is the bar chart of academic use, which rises steeply for
several years, then is flat for a few.  The authors explain the flatness
by academia's resistance to change - never mind the initial steep part
of the graph!  It seems to be quite full of bizarre statistics,
projections, and comparisons (Ada gets a score of 5/5 on type safety,
C++ a score of 4!)  And, as pointed out, it seems reliant on the (rather
old) web site  But at the end the editors do say a
future article will offer an opposing view.  I'm curious to see that.