While this setup is awesome, and I have enough DSPs and CWA/3rd Party plugs to not have much use for any VSTs (maybe one or two), I really want to take more advantage of Warp's newer verbs, use more instances of SpaceF's Echo, etc. but I keep hitting the PCI limitations of my laptop.
I am going to try and see if getting a PCI interface for the Magma (as opposed to the Cardbus interface) and putting it into the slot in my laptop's dock will give me any advantage. While this introduces an extra bus to pass through, my thought is, the dock's ports probably are given higher priority, and may put the CWA cards on a different IRQ have better bandwidth than the cardbus is giving me.
Right now, the Cardbus/Magma port is shared with USB, my onboard video, the onboard sound, and the ethernet port. If this doesn't work, I will try something else, which is really the subject of this post.
I'm getting a Shuttle mini-low-power thingee with an nForce 3 Socket 939 mobo for another project. It's got one PCI slot, which would be enough for the Magma PCI chassis to utilise. I want to try and see if this would make a useful dedicated CWA box, then get an external audio card, be it USB, Firewire, or Cardbus that has only ADAT on it (or if it has more, it's not going to alter the price on what I hope would be decent quality for kinda cheap).
I've looked at the RME options, but they're SO DAMNED EXPENSIVE!!!! ...and you don't even get any plugins.

If Shuttle hasn't altered the nvidia reference design in such a way that harms SFP performance, then I can FX-Teleport the two machines together, and use TightVNC to handle display the CWA information to my laptop or it's 2nd display.
Also, using TripleLE, or VDAT, I could use the laptop wirelessly to track in an isolated room, or far away from my I/O's without having to get up, or introduce a lot of traditional computer noise near the mic's range.
Any ideas? Recommendations?