Substring match is important to us, so we'd love to see Squat reinstated with speed improvements. It seems like Solr can handle substrings as well ([Edge]NGramFilterFactory), but for small deployments, having the engine built right in is a plus.