In this case, each micro tree has 80 nodes, or 240 channels. Lets say you have 17 of these that you want to run in your display (240*17<4096). Normally you'd need 2 Hubs to do this. With the passive hub you can run pixel net to the main hub, output to 15 SSRs and to the passive hub, then output from the passive hub to 2 more SSRs. You now have 17 elements and only had to spend another $8 for a passive hub rather than $65 for another main hub plus more for another ATX. You can also run it all off the single power supply (assuming you have the amperage), you just need to shunt an spt2 line alongside the cat5 running from the ATX/main hub enclosure to the passive hub.
Each bullb on every micro tree (or whatever) can still be programmed individually. Remember, the bulb channel numbers are set in the SSRs, not the hubs.
You would use the passive hub anywhere you wanted more than 16 individual items in your display that were separated by an inconvenient distance to run as single RGB node string to multiple elements. This keeps you from having to splice in extra cable between micro tree 1 and 2 for instance in order to have the bulbs on both trees attached to a single line. The down side of this, is that in addition to the passive hub, and power line, you need an additional SSR to power the new string which cuts back on your savings.
I'd only add this. To avoid confusing somebody make sure we're referring to the controllers by their correct acronym. Smart Strings use
SSCs - (Smart String Controllers)
NOT SSRs - (Solid State Relays). SSRs are not compatible with any of the smart string hubs, etc.
Call me picky but it can mess people up if we mix up our TLAs (Three Letter Abbreviations).
Thanks
Alan T