2^16 rounds of SHA-256 might not be "ultra slow" but it's certainly not as bad as you make it seem. If you read the analysis in the article, it would take many years to crack a random password with this hash function. It's unfair to say I "ignored" that requirement.