[chavesn]

All of that statistical analysis was actually a bit silly, because I've never heard the "typo" argument as a reason for email grammar validation[1]. Sounds like a straw man. It didn't need to be disproven.

The conclusion is sound (although leaves out a discussion of the whether an email confirmation field is at least better than nothing).

[1]: (As a side note, I think the most common explanations for grammar validation are programmer perfectionism and proactively stopping user garbage, such as copy-paste errors or intentionally fluffed fields that will result in a bounced email anyway.)