The whole point of George Turner's article is that COP has nothing to do with blade vibrations, exactly opposite of the "harmonic balance" theory. As far as vibrations are concerned, your sword hand location should be close to the handle node, to minimize felt vibration no matter which part of the blade hits the target - this is what I would call a harmonically balanced sword (plus, maybe, the blade node being close to the true CoP). COP is calculated from mass and geometry of all components of a sword based on their moments of inertia, or (much easier) is measured experimentally by the pendulum method. Having a theoretical equation is useful when you have to compare swords that you can not handle personaly - e.g. when shopping on-line! And the sword's moment of inertia is a much more important factor than its COP, since MOI tells you how easy it will be to accelerate/stop rotational motion of a sword, in other words how a sword would handle.
I have worked out a more-or-less simple model to calculate MOI of a tapered sword, but it requires a lot of informatoin to be useful - like degrees of profile and distal taper, exact mass and dimentions, and a good idea about relative masses of the sword's handle, pommel, guard and blade. Very recently, I have asked at the SFI if anyone would be interested in it, and got exactly zero responses

I thought everyone would want a tool to estimate a sword's dynamic balance without picking it up, shows how much I know

At least I did not get flamed for blasphemy.
Regardless, here it is: (x**2 means x squared)
I=2(b**2)*M
tot + (a**2)(M
p + (1/3)M
h) + a*b*(2M
p + M
h) + M
b[((L**2)/2)-L*b] + M
b{(L**2)[((t
1*t
2)-5)/(30*n)] - L*b[((t
1*t
2) -(t
1+t
2))/(6*n)]}
where
n=1+ t1*t2*(1/3) - (t1+t2)*(1/2)
a = handle length
b = distance from pivot point to Point of Balance, or Center of Mass (use the usual PoB-to guard distance to get I relative to cross)
L = blade length
Masses: M
tot=total, M
h=handle, M
p=pommel,M
b=blade (tang not included, it's part of M
h, sorry).
t1 and t2 are distal and profile taper, linear approximation. For example, no taper means t=0, taper to a point - t=1, t=0.5 for 50% taper, etc.
I'd be glad if people tested it, seems to work with my repro saber.
CoP is easily calculated from this, just find Icom=I - (b**2)Mtot, and then for a pivot point (somewhere in the middle of your sword hand for a single hander) at distance X from CoB (balance point), your CoP is on the other side of CoB at this distance from it: Icom/(X*Mtot)
Enjoy!