Hi,  Here is my view on the subject.

Let us look at a ice skater turning around his own axis with arms
stretched out sideways.  He accelerates initially and the spins around
at a certain speed.

Compare his one arm being a hammer shank travelling.  Now, if he pulls
his arms in or bends them, he will spin faster because he concentrates
the mass more towards the centre.  Likewise, he will spin slower again
when he extends the arms.

Now, if the hammer shank does bend, the hammer must in effect move
closer to the centre pin (unless the shank would stretch => unlikely).
This means that, in effect, the mass is more towards the centre at the
time of acceleration.  Once the hammer is in free-fall after letoff,
the shank will straighten, move the hammer further away from the centre
pin, and as a result, it will decelerate (arm stretched out further).

Some other point to note here is that the energy transmitted from the
hammer to the string is constant, flex or no flex of shank, because the
moving mass does not change.  I therefore say that it will not
influence the volume of sound at all if the shank bends a little or
not.

Maybe someone will come up with a theory that the hammer's mass changes
after letoff.  This would be very interesting to read. :-)

Bernt Damm

 [ Although the center of mass (hence the inertia) might shift slightly
 [ as you describe, I think the energy stored in the compliant shank
 [ dominates the situation.  Think of a limber golf club, or the pole-
 [ vaulter with a fiberglass pole.   -- Robbie