Yes. The crap people publish even with peer-reviews is bad enough. A few subfields in CS have bitten a couple times by relying on non peer reviewed ecosystems.