Well, now the results of the models in Inventor, also confirmatory of slight benefitial action. Also the table of comparison of results, updated. Those signaled in ellipse are the first ones that signaled what is becoming an statistic anomaly and I will try still to repair if feasible to commonality.
Thankyou to everyone for pushing the clarification so far.
And the reconciled Algor Simulation models. I took no chances varying a number of things whilst preserving the shape and intent of the models and quite likely the use of a finer mesh (even if the former was a fine one) did. So, friends, all my pairs of FEM models say now that the addition of stiffeners signify a slight improvement against rotation.