P51 Oops

You've now got my brain hurting Huh!

Here's a few thoughts:

1. The decalage should be irrelevant, it's a radio model so you set it with the elevator position in flight.

2. It's either unstable due to the C of G or something's moving that shouldn't.

3. C of G, as Eric says, balance it slightly behind the wheels and see what happens.

4. Assemble the model and check the security of the wing, tailplane, ailerons, flaps, elevators.

5. With an assembled model, set up elevator, get someone to hold the model at the wing root, then try to gently bend the elevator back to neutral. Does the fuselage bend, does the tailplane twist, do the control rods distort, etc, etc.

Can't think of anything else at the mo Klippy, but if I do, I'll post it. Notwithstanding, good luck with the model.

As you have suggested a 'floppy' rear fuselage is likely to make the elevator control response floppy as well rather than over sensitive.
This can be an issue if you get into a screaming dive as in an extreme case the fuselage flex can completely negate the effect of the elevator and its lawn dart time!
Of more concern is if the fuselage is flexible enough to 'flutter' which is likely to lead to a structural failure of something.

Allways a bit entertaining Klippy.
Seems more than half the guys I fly with in various areas of Scotland are older than you, including me.
Small control thows. Expo IS very useful. A gentle, level chuck it, and see what it does ?? without using radio tx.
Fingers crossed for you