It depends strongly on the subject, I find. Sometimes this advice is right; sometimes it's very far off. I've made both mistakes, and it's hard to predict which is which. Sometimes you read forever when you should've just done something months ago. But other times you waste a lot of time badly reinventing a wheel and getting saddled with bad design decisions that should've been avoidable, because you didn't realize that what you were trying to do already had a name and a lot of smart things written about it (or realized but didn't want to read them). You might call that the "just get down to business and parse HTML with a regex" school of getting-things-done, to take an extreme but common example that might've benefitted from more reading. ;-) A lot of bad statistics are also in that category...

I think the 2nd mistake might actually be more common, given the amount of wheel-reinventing (and not always reinventing well) that goes on, but it's hard to say.