I would add that there are relevant nuances that were not discussed in the Chess Life article. For instance, the strongest line against the Modern Benoni is the Taimanov Variation, which includes playing f4 before Nf3. If black plays a Nimzo move order, 1 d4 Nf6 2 c4 e6 and white avoids the Nimzo with 3 Nf3, then the f pawn is blocked from playing an early f4, and white cannot play the Taimanov variation. Black can transpose to the Modern Benoni with 3 ... c5, having avoided white's strongest line. The article did not assess the strength of the Modern Benoni conditioned on avoiding the Taimanov line.
This is just one of many examples where assessments of openings are very nuanced by move orders and transpositions, so that one really has to assess an overall systematic repertoire with the web of transpositions possible with the move orders used to get an overall assessment of the strength of the openings.
GM Larry Kaufman has a 2-part series on engines in Chess Life. Aggregating data from top engines, he provided their algorithmic assessments of openings. This does not necessarily mean the ones they rated the best are the strongest, but it should not be dismissed either. His presentation was that the King's Indian, Dutch, Modern Benoni, Benko Gambit, and Budapest are all rated as dubious or borderline dubious. The QGD Orthodox Defense, Slav Defense, Queen's Gambit Accepted and Nimzo-Indian were rated as the strongest replies. The Gruenfeld was rated only a little less solid than those. The Queen's Indian was rated a bit lower-- not in dubious territory, but if white plays 3. Nf3 to avoid the Nimzo-Indian, then 3 ... d5 transposing to a QGD Orthodox Defense would be rated stronger than a Queen's Indian.
Of course, at the amateur level, all of the above are playable. Nonetheless, if you are playing someone USCF 2200+, you may have a rougher time with some of the less solid choices.