Chris
The major disadvatage of your approach is that you lose the ability to dim mimics. This is a key issue in bedrooms for example where the mimics need to be kept very low at night such that they can still be seen but not bright enough to light a room.
There are various laudable approaches to reducing the number of WebBricks needed for an installation, however these always lead to higher costs and lower profits in the long run. I'll outline some of the important issues here:
WebBricks are good value compared to installer time, therefore it makes sense not to over populate them and give yourself cabling issues. It also leaves room for expansion and copes with te inevitable 'changes of mind' at the customer site. So here are some rules of thumb:
a) Try to avoid remapping mimics, it takes time and concentration! perhaps the only acceptable route is to map onto the 'lead' analogue channel in scenes, personally I wouldn't do it because you have to remember this stuff.
b) Temperature sensors should be used for local control wherever possible - if you rely on the gateway all the time a customer induced network problem could cause a cold shower -- customers don't like this.
c) Use the fact that the Gateway (HGA) has multiple Ethernet interfaces, these can be used to separate the WebBricks from the general LAN - thus increasing reliabilty.
d) If you have got enough outputs for a particular task, tie the inputs of one or more WebBricks together instead of relying on Gateway(HGA) programming -- its quicker and more effecting, you'll get more mimic channels that you can shake a stick at.
e) Use mimics as mimics, we all know they'll drive a pile of other stuff but the debugging effort is not worth it.
Mimics will be supported as mimics, fault finding configurations for mapped or re-purposed mimics may incur support charges.
Regards
Andy
PS I know this sounds tough, but all our reputations are on the line here, repeatable reliable and understandable installations should be our common goal.