There are three primary reasons to add a second sub:
1) To increase headroom (up to 6dB if co-located);
2) To smooth frequency response (reduce room nulls and nodes);
3) To reduce localization (if that is an issue).
I suppose there is a fourth: It feels cool!
(That one works for me, anyway.
)
Placement depends on your priorities. Co-locating them will maximize headroom, but won't do much for smoothing or localization. Separating them is more likely to help smooth the FR, but will probably not increase headroom by as much. It's also more challenging to find the right placement, phase relationships and the like.
Conventional wisdom is to use two of the same model, but my experience is different. It would probably be more important if they are co-located, but I've always separated them since my priority has been to smooth FR. (I have a 16-46 PC+ together with a PB12-NSD, and they work well together.)
If you "need" more output, a PB12-NSD would match up well. It's tuned similarly, but delivers 3-4 dB more than the PB10.
As I understand it, there is not much, if any, difference between the PB10-ISD and the PB10-NSD. In their other subs, the new drivers deliver 1-2dB more output than their predecessors, but I don't believe that holds for the PB10.
I have no specific knowledge of the used market, but SVS subs hold their value, so I would guess the PB10 would return 2/3 of its original value on the used market.
But my primary recommendation to you would be to ask these questions directly of SVS. Ed or Ron will be very helpful, and know more about some of these comparisons between their different models than anyone here can.