Not sure how far along in development a next-gen Virus is, or if development has even started, but I have a few suggestions on connectivity.
We know USB 1.1 is an atrocious medium to send audio to and from the Virus, especially in 2012 when most chipsets are designed to take advantage of newer high speed connection protocols.
USB3.0 has now been out for a while and looks to be the standard for some time to come. However, with the problems people have been having with their particular USB environments and current-gen Virus integration, I thought taking a more generalized "cover all your bases" approach to connectivity would be a huge step forward for Access in building a new synth.
Take MOTU for example, who a couple years ago pioneered the first ever hybrid Firewire/USB devices. Not much of a technological breakthrough, but tremendously useful for someone trying to maximize performance in a highly variable (from one studio to another) environment. If USB is a no-go, simply switch to FW. If you don't have FW ports, use USB until you can add a card. If FW gets reduced performance in Windows, use USB. If your mac book only has 2 USB ports, use FW, and so on.
Thunderbolt is promising as well, and will be coming to PCs in the near future. It has insanely high data transfer speeds that leave USB and FW in the dust. Maybe this is overkill for a digital synth, but it might be worth looking into.
But my revolutionary suggestion, which I'm not sure has been suggested yet, is to add a PCI(e) card to the mix. Having an Access-developed dedicated card in the motherboard could do wonders for latency, hiccups and other glitches so many have been experiencing. Instead of messing about with hubs and ports (which you are unsure if they are on hubs or dedicated ports at all), why not simply connect the next-gen Virus to a dedicated Virus PCI(e) card via whatever connection protocol would work best?
Why not make such a card an optional add-on for those that want to take Total Integration in their studio to a new level? Maybe Access could even design something specifically for the current Virus TI and begin testing it. So many people have resorted to third-party chips to add USB ports with little to no success in bringing about a fix to their issues.
I guess my main point would be relying on any one type of connection has, in the case of the current-gen Virus, proven a massive headache as technology progresses and the Virus is stuck at USB 1.1.
In my opinion limiting connectivity for a next-gen synth to save time in software/firmware development would be a huge mistake. Limiting connectivity to make the synth more economical would be as big of a mistake; people are shelling out money for a working product for their studio and are happily doing so. If it is proven to work in all circumstances, then they will feel it is money well spend and well worth it.
Yes, adding FW, TB and direct PCIe connection over some other protocol would increase time spent developing the firmware to levels I'm not sure they are comfortable with yet. But I feel this is a necessary step if Access wants to sell a product that anyone can use, out of the box, in any system, with no hiccups or headaches, and isn't that what TI is all about?
What we have now is a synth that claims total integration but isn't totally integrated unless it is in the perfect environment that isn't quite possible to reproduce, and isn't fair to consumers/musicians who have environments that vary drastically. What we have is a fish, which is totally integrated for an aquatic environment, but will die on land. We need an amphibian. We need a mutation.