To clarify, I maintain that testimonials from people who give their full names (and thus can be contacted after) are perceived as solid evidence. A common real-life example of named testimonials seen as credible by recruiters are written LinkedIn endorsements from named people who are connections on a person's profile. However, I agree that nameless or semi-anonymized testimonials are less valuable and give the entire term of "testimonial" a poor reputation, because their truthfulness can't be verified.

On the second point, what you wrote is true, but the Airbnb mention was pretty short; your comment is probably around the same length as Graham's mention. The Airbnb mention in full consists of: "[4] The Airbnbs were particularly good at listening — partly because they were flexible and disciplined, but also because they'd had such a rough time during the preceding year. They were ready to listen."

I could find no mentions of other named companies involved with YC in the article, and the Airbnb mention was quite brief (the assertion was that they listened to YC's advice, and the implication is that this was the reason behind its success).