Two things have always bothered me about this paper:|
1. It was written after Modula-2 was developed, a language which certainly fixed many or most of Kernighan's criticisms, so obviously even Wirth himself knew that the language was not suitable for many uses. It seemed overly pedantic to point this out after the fact.
2. It only took some minor additions and extensions to turn Pascal into something that was found useful for industrial strength programming. Witness Turbo Pascal and VAX Pascal, for example. Kernighan's paper could have taken the angle of "What Pascal Needs to Become a Very Useful Language," which would have come across as less angry.