Modified the default maximal Eddington fraction in all the EAGLE example to 150% to reflect the mistake from the original EAGLE runs where the fraction was set to 100% / h.

Implement the reduction of the accretion rate of the BHs based on the local viscuous time-scale in the EAGLE model.

Changed the way we set the run frequency to use a physical simulation time rather than a fixed number of steps.

Use a better value (the default one) for the gas fraction assumed by the EAGLE SF model in the EAGLE low-z examples.

Make all the EAGLE examples run with gravity, cosmology and stars switched on. Make them dump more snapshots.

Set the compression level to 1 for the basic examples. 4 does not improve things and is just slower.

Improve the way we define the time of the first snapshot and statistics in cosmology and non-cosmology modes.