Umm...no, not normally. It's a factor of center of gravity vs center of lift. If I have to push down on the tail to keep the nose of the plane up, and then I lose that push, the nose goes down. While the math is a bit messy, the concept is not.
Agreed, at least that's what we learned in Aero Eng.
What they're seeing is probably a dynamic effect. Modelled in level flight, I can't imagine a scenario where the airplane wouldn't pitch nose down after losing the horistab. THat's just because, in statically stable aircraft, anyway, the cg is forward of the Center of Pressure - the extent of which is labelled the static margin.
However, IRL, if you're already pitched nose down, you're doing so by creating lift on the elevator. Lose that and the nose will likely pop up.
Of course, even with the stab locked and power off, most a/c will exhibit classic phugoid (long period) motion, nosing up until lift abates from dropping KE, dropping the nose until lift recovers on both surfaces and causes pitch up. That's the nature of longitudinally stable a/c - self-recovering divergence. The short period motion has period proportional to the static margin to which you refer earlier (diff b/w NP and CG).
All I'm saying here, short form, is it depends on the force being created by the horistab at time of "separation event".