I don't think most ML people are actively hostile to HTM, just indifferent until they show some results. For example, Yann Lecun in his Reddit AMA: http://www.reddit.com/r/MachineLearning/comments/25lnbt/ama_...