I don't know nothing about DAS, but it is logicall that the passive DAS just helps Shannon capacity on the uplink and not the downlink because you still have to blast the data on all the network antennas to get to the phone.  Now if you have active DAS and only transmit downlink data to the specific antena that is closest to the cell phone, then all you are doing is solving the Shannon problem but you still have serious capacity problem trying to address all of that data through one base station--especially for LTE.  It looks to me that DAS is a bandade.  These telco dinasaurs will eventually have to add more base stations to keep up with the Cooper's law.  

Interesting point. I do wonder to what extent femtos/picos/indoor small cells have put a dent in the DAS market -- have they have a big impact?

I'd like to point out that femto cell solutions like these have numerous drawbacks that are addressed by Distributed Antenna Systems (DAS), which is why DAS is the most widely used in-building coverage and capacity solution. Unlike femtocells, DAS support 2G, 3G, and 4G protocols simultaneously and also support multiple frequencies to handle all of the flavors of mobile service one or more operators will provide. In addition, DAS automatically takes care of soft hand-off issues that arise when individual femtocells simulcast a signal inside a building.