I have to answer in a way that is different than the questions to keep from being misleading.
The use of a named server in the OSD, a variable in the OSD, or the use of ASR in a client registry, are simply three different ways to direct the client to a server when an applicaion is launched. If ASR is present, it trumps whatever is in the OSD, variable or not.
When a user attempts to launch an application and the client is directed to a streaming server rather than the full management server, the launch is executed without the possibility of added licensing enforcement. Of course publishing only made the app available to authorized users anyway, but you are asking about licensing because you wanted more.
That the launch occurred (including start/stop times and result/return code by the app) is stored at the client in an XML file stored in an obscure area. The next time the client performs an application refresh operation using RTSP/RTSPS with the full management server, those results will be uploaded over the RTSP/RTSPS channel and recorded in the database for reporting purposes. This mechanism also improved the reporting for notebook clients that run apps offline (prior to 4.5 offline reporting was not recorded).
So with streaming server, that is as good as it gets. If a remote site uses App-V 4.6 clients on only Windows 7 or Windows Server 2008 R2 servers, then skipping the streaming server and using Windows Branch Cache becomes a possibility. The remote site clients get the full features, but the bulk sft/osd/ico transfers will get cached to reduce bandwidth (and latency) while authentication/authorization traffic will go all the way back to the main site server.
tim (Microsoft MVP for App-V)
Kahuna, TMurgent Technologies: http://www.tmurgent.com
President, Virtualization Boston: http://www.virtg.com