Huh? From your article:
I should say right away, I haven't found anything crucial and I think there are three reasons why PVS-Studio hasn't found serious bugs
The rest of the article doesn't seem to back up your claim really. It even says "The code is of high quality."
Regardless I agree the article isn't very good, serious or otherwise. If it is serious it uses one anecdotal piece to make a sweeping claim on the productivity of programmers across nations...
They show bugs in cfront, trying to sell their product to C++ developers. They'll likely call cfront high quality.
That said, Roskind's experience is more telling than the output of a static analyzer because generally and certainly with a compiler, actual testing shows orders of magnitude more issues than static analysis.
Since I'm not sure that this was written seriously, I'll point to anecdotal evidence of cfront, Stroustrup's C++ compiler, being the buggiest C++ compiler ever written. (Here's a bunch of bugs found by modern static analysis tools - http://www.i-programmer.info/programming/cc/9212-finding-bug... - but unfortunately I can't find the Usenet old posts telling that cfront was the buggiest C++ compiler of its time, and that early g++ was the next buggiest. The Unix-Haters Handbook has evidence from James Roskind who tried to make a yacc grammar for C++ - to a first approximation, an impossible thing to do due to a few goofy design decisions in the C++ grammar - and who said that every time when he tried to feed cfront input that would enlighten him with respect to some dark corner of the grammar, cfront crashed.)