I believe if you have any type of coffee while buying on Amazon, you have to pay them for that too ;)

difficult to know how factually accurate the book will be

Good point. I remember when I was learning Perl (in '94) there wasn't the saturation of the market with Perl books like there is now. So, you had to trust what was there. And even today, without a person researching who an author is, or who has given good reviews of a book, it is hard for them to know what to trust. For example, the name Randal Schwartz doesn't carry much value with someone who knows nothing about Perl, unless they research the name...

So, people hopefully do their research to make sure the author of the book knows what they are talking about, or choose what reads best to them.

I have had some conversations in my Perl Monger group about how many people learned something about Perl by using/reading/hacking on Matt Wrights scripts. We all now know that they are bad, but at the time of Perl 4, they seemed like good examples to many people. Many people learned bad things, and some unlearned those bad things.
It's the same with books. If you know nothing, and pick up a bad book, you will likely learn some bad things, and later need to unlearn them. If someone learns with Perl for Dummies (a bad book), they will hopefully learn enough about Perl itself that unlearning bad programming practice will come with reading other books, and seeing code, and participating in the Perl community. It takes a community to raise a good Perl hacker! :)

Cheers,
KM