He didn't seem to be arguing against the principle of charging for the service. But, from a user's viewpoint, having to pay for something is rarely an argument in its favor.
It's called "trying to capture more value than you add." Seems to be a key idea behind many SaaS/PaaS startups these days. Not an indicator for long-term success.
Well that's a judgement call for the user choosing the service. If they think the value is there, they will pay. That's how capitalism works, and I daresay if Gitbook had it wrong, they'd have given up by now.
What experience do you have with Gitbook to deem that it doesn't add enough value to deserve the payment?