Questions on the enigmatic rotational curve of spiral galaxies

I just want to add a thought experiment to help us visualise what is being
suggested/asked.

Consider a number of stars arrayed in a line across space, say a few light
years in length.

Assuming the initial condition is one of a stationary motion of each star
relative to the other, what would we expect to occur next?

I assume that the mutual attraction of the stars will cause the row of stars
to shorten until they end up clumped together.

So to make my 'barred' galaxy stable I rotate it fast enough so that the
stars on the two ends don't proceed toward or away from the rotational
centre.

Will this work? I assume it will not work close to the rotational centre
but will work further out.

There is a greater gravitational pull on objects closer to the two ends
because there is more mass between those objects and the centre, but the
rotational speed is greater as well - nicely balanced

Now we add more mass to the centre. Objects toward the ends of the arms are
going to be drawn inwardly unless the arm describes an arc. Now the pull
directly from the central mass can be added to the less effective pull of
the curved arm and the galaxy is again stable.

Why do I think this has been missed?
Models consist of known and unknown quantities, such as the numbers of and
masses of the stars that can be observed. But unobserved is the central
mass which is estimated. The central mass, I assume, is estimated at far
higher than it actually is. I don't think modellers have considered
relatively tiny central masses, as in my barred galaxy model above. Thus a
higher central mass then requires a greater mass beyond the galaxy for it to
rotate as observed.

I do not have the skills to test these ideas but I assume that either:
It has already been considered and shown to be flawed/viable or
members of this list can estimate the viability of the idea using a
simplified model (as I have suggested above).

Robert