This is not a benchmark of some model on cifar-10, it's a benchmark of the training algorithm. But, model size and complexity also matters. MLP with backprop g

This is not a benchmark of some model on cifar-10, it's a benchmark of the training algorithm. But, model size and complexity also matters. MLP with backprop gets about 63% on cifar-10, for various reasons. So achieving 59% accuracy means this algorithm is about 93% as good as backprop in this case. However, 63% accuracy on cifar-10 can be achieved with two (maybe three) layers IIRC. The output is a 10-way classifier, which is handled in one layer. If the output requires multi-layer transformations, then gradients need to be back-propagated. As long as the batch activation vectors are trained to max separation (or orthogonality or whatever) at each layer, one output layer can match them to labels. But this is unlikely in problems where the output is more "transformed or complicated".