Nathan has pretty much hit the nail on the head. And the nuance that Oliver has pointed out really just emphasizes Nathan's overarching point that "I don't think it's possible to be definitive".
Most of the arguments I've been in with people about these types of design decisions are usually rooted in debate and not in the interest of good design possibilities. The more you design (for any medium), the more you'll learn that there is no right or wrong. There's just really good or really bad ;). And in many cases, like Nathan has mentioned, people are interacting successfully with a number of different design solutions and patterns. This should not be a surprise when we consider design in other aspects of our life aside from computers and software. For instance, some countries drive on the right, some on the left, and some on both.
In the case of web forms, the important thing is that you're consistent and that you exercise good visual design principles. So it feels like if fits with the over arching aesthetic, general content layout, spacing, and color. The basics can go a long way if done well.
Sure, Nielsen's research is useful like any other research (it's nice to have samples). But if the heat maps were measuring a form with a big huge red button in the bottom right of the page, all of a sudden the E-pattern is the new F-pattern if you get my drift.