I believe you can trust the directly quoted stuff, keeping in mind that people remember things differently. Anything where the author is paraphrasing without citing identifiable sources requires outside confirmation to be trusted. At least, that's the case for someone like me, who does a lot of research and understands all too well the fallibility of the researcher/author. I never just take one guy's word for something. It's nothing personal against Eliot; it's my practice in general because even the most well-intentioned person makes mistakes.