Training or Punishing?

Someone sent this article to me via email:
http://www.seespotrunkennel.com/blog/punishing-out-warning-signals-reality-or-myth/
I just felt I had to address some of the author's points. She argues that correcting a growling dog does not always create a dog who bites without warning and that it's only anecdotal evidence that is behind this thought process.
She also points out that people like Pamela Dennison and Pat Miller, (and even Ian Dunbar); people who encourage more positive methods of training are the people who believe and promote the idea that correcting a growling dog creates a stealth biter.
She does make a great point that growling is only one of the first signals that dogs give. I have pointed out that freezing, baring teeth, a hard stare, and a puckered commissure are all signals as well.
I then ask: Will the average pet owner be aware enough to watch for the other signals? Or will they say "He bit me without any warning?"
Ruth also states that a dog who growls at you for trying to take his bone is stating "Just try to take my bone, b!+@#" or words to that effect.  This statement smacks of the anthropomorphizing of dog trainers from the 1950’s and 60’s (that all dogs are stubborn and resent being trained).  Her argument is that if punishment is well timed, the dog can be taught what is appropriate and what is not.
I'm sure this is true. I'm not a positive-only trainer and I do use corrections. In my opinion, the flaw in her argument is the average owner, even the average trainer, has very bad timing. Whether using a clicker or a collar pop most people don't have the experience or skill to capture or correct a behavior.
So I pose the question using her scenario: What's wrong with training the dog to allow me to take the bone, instead of just punishing the growl?