It's not a bad school by any stretch, but it's not on the same level. Looking at federal grant money (e.g. NSF, reasonable metric for research), U Pitt is at 26 million to CMU's 70 million. Penn is a much closer match at 48 million. Berkeley and Stanford are at 112 and 78 million, for comparison.
You can make 'world class' mean whatever you'd like, but I think most would consider Pitt a solid state-level school, whereas CMU is more competitive nationally.
I'm definitely not saying it doesn't matter, but it's not as significant as CMU, and PG's argument seems to be that you benefit from having a really apex research institution for making a startup scene.
You are correct that CMU and Pitt are not on the same level, but you have your dirctionality flip. Pitt is in top 10 in research expenditures nationally[1], I don't think CMU is even in the top 50. Don't ignore the significance of a medical school. NSF has a 7 billion dollar budget, but NIH is 30 billion. Don't even get me started on endowments.
Pitt+CMU == complementary research institutions that have had a long tradition of collaborating (for example the Pittsburgh Supercomputing Center) and BOTH provide strong intellectual anchors for the city. BOTH are significant.
NSF funding isn't really a fair comparison, because Pitt is most well known for its excellent medical research. $4.5 million in NIH funding at CMU vs $79.9 million at Pitt in FY2016...
Plus CMU and Pitt collaborate quite a lot, especially in biomedical research.
You can make 'world class' mean whatever you'd like, but I think most would consider Pitt a solid state-level school, whereas CMU is more competitive nationally.
http://dellweb.bfa.nsf.gov/awdlst2/default.asp
I'm definitely not saying it doesn't matter, but it's not as significant as CMU, and PG's argument seems to be that you benefit from having a really apex research institution for making a startup scene.