1)In Ephemeris Mode the state vector output for min date and max date of
integrator is different if it considered the Bounded propagator or
not.Are this 2 points in the set of bounded points?
2)It would be better if the orbit, in ephemeris mode, were represented by Analitic formula(parametric order and type polynomial );maybe is more efficiently in time for example if you want retrive often different single point or if you want store the calculated orbit in file and use it later, and in this way you would have separate the orbit propagation with other utility (orbital event ecc).
3) Droziner an Cunninghan models have a problem if high order coefficients are considered.
(from redmine: issue id 14, created on 2011-03-07, closed on 2011-03-07)