I'm not 100% sure I understand what you're asking for, but since you'll never fill the box 100%, does a few GB's difference in you calcs make that much difference?
You can look at the free raw and space for a given class of disk "showsys -space -devtype FC" and look at the free initialized space, apply your overhead e.g 25% for 3+1 and then divide by 1024 and you have usable GiB's. You could do the same for total capacity rather than just free.
showsys -space -devtype FC
Free (15138816 * 0.75) = 11354112 / 1024 = 11088 GiB
Total (23396352 *0.75) = 17547264 / 1024 = 17136 GiB
Alternatively you could use "showspace cpgname" to forecast amount of free space available to a given CPG's or "showspace -t r5" to play what if, similar results but a bit more exact as it takes into account your CPG layout.
showspace -t r5 -ha cage -p -devtype FC
HA Cage (15118336 * 0.75) = 11338752 / 1024 = 11073 GiB
showspace -t r5 -ha mag -p -devtype FC
HA Mag (15130624 * 0.75) = 11347968 / 1024 = 11082 GiB
You could also work this backward based on the current CPG utilization. Failing that I'd look at writing a perl script or similar to pull the various numbers and provide the output you need.
I noticed to mentioned Admin space, unfortunately this isn't a fixed size and will depend on how you are using the system. But think of it this way, if you don't set aside admin space for metadata you don't get all the other 3PAR benefits.
Just my opinion but on a traditional array you know exactly to the GB how much space you have at any point in time because you've pre-configured and pre-comitted the space upfront. Get it wrong and you've wasted lots of space typically requiring a data migration or more usually additional disk to put it right. Then again who writes to 100% so knowing the exact to the GB is not really that useful from a practical point of view.
With 3PAR you don't commit things upfront and you can just change things on the fly, metadata gets you flexibility, so don't think about it as wasted or unusable space.
http://h20195.www2.hp.com/v2/GetPDF.asp ... 987ENW.pdf