And we could, if we were bothered, add that we don't know* where* life started on the Earth. If it were deep in the oceans at volcanic vents, as is possible, then the presence (or not) of an atmosphere or a magnetic field makes absolutely no difference to the basic chemistry.

Correct. And we could also remember that a month ago, initial simulations suggesting that migrating HJs would make earth-like planets impossible turned out to be an exaggeration. Simulations can be, and often are, way off beam.

The paper arXiv:1010.5133v1 [cond-mat.mtrl-sci]

http://arxiv.org/abs/1010.5133 comments that we know nothing about the iron melting curve under conditions corresponding to planets of several times Earth's mass. This is beyond our experimental capability. So they have performed theoretical calculations at the atomic level, and note that higher pressures make iron solidify at higher temperatures. They then match this theoretical curve to another theoretical curve for planetary adiabatic cooling, making assumptions about thermal barriers, and come up with a suggestion that cores will solidify quite rapidly for planets of 2xE and above. There is a lot of theory there, and a lot of assumptions.

They finally comment that precise conditions are, of course, unknown - temperatures will depend on the age of the planet and the precise alloy mix of the iron will affect the melting point. Finally, they comment that convection in a salty ocean may also generate a magnetic field, a suggestion which has been made for the one on Ganymede.

So the paper is highly theoretical, and many steps from saying that large earths will not have magnetic fields....