@YOUSEFY Thanks for your (promised) reply. I am glad you wrote it. After studying material on parallel algorithms recently, I also started to wonder why field of "algorithm" is not focusing on "parallel algorithms". Or rather, I wonder why there have been bursts of interests, like between 1987-1991, followed by long periods of neglect. Mismatch between models and hardware might have been one factor.
While I well know Frobenius norm, thinking about this graph isomorphism problem, I have accidentally re-invented Frobenius inner product: kind of scalar product on matrices, which is rotation-invariant:
Rotation (change of basis) is a natural matrix transformation (let's focus on real $\mathbb{R}^{n\times n}$ here):
$$r_O(A)=O^TAO\qquad \textrm{for some orthogonal}\qquad O^TO=OO^T=1$$
There are well known 1-matrix operations which are rotation invariant - trace of powers and characteristic poly...