EAGLE SF should use the actuall pressure not the EoS one

The pressure used in the EAGLE SF should be the actual pressure of the particles and not the pressure on the polytropic EoS as is currently done. This means that gas with P > P_{eos} (but still within 0.5dex of the EoS) should have a slightly higher SFR.

I misrunderstood the original paper and I think told you that P was always computed from the EoS but that's incorrect; at least in the actual EAGLE code. This will mainly have consequences when the EoS is lowered or removed.