From the website text:
My own interest in such phenomena started with the discovery that 80% of scientific citations are not read by the citing authors but copied from the lists of references used in other papers
I question this. I am a scientist myself and I NEVER quote papers I have not read.
I do know however, that when compiling your reference list for final submission (always the last chore to do), it is sometimes easier to quickly Google for the volume and page numbers or lift them from review papers than to check the actual papers again individually. That does not mean however that the papers in question were not actually read earlier in the process of conducting the research and writing the paper. It does explain how misprints get perpetuated in reference lists.