If the hack works as it should there will be no side effects.
The PCIE format works the same across all motherboards. If you have two 16x PCIE then theoretically you can put two of the same PCIE cards in them (either Radeons or Geforce cards) and then link them. Sadly due to monopolisation it's locked out.
For example on a Crossfire motherboard the bios includes an instance. When you get into Windows your chipset is then recognised so that the drivers can look at it. The equation is that the drivers look to the chipset identified by Windows. A simple program line of (not in C) "If ID = XXXX allow". So, ID being the hardware ID given by the GPU and the XXXX bit being the specific hardware ID used by say a Radeon. So, if an AMD chipset is Crossfire only (which they are, due to them being Nvidia's biggest rival) then the drivers will look at the ID of the cards and refuse to link them.
SLI works in the same way. When you boot into Windows it looks at the cards. However, the protection to not running Crossfire on an Nvidia chipset (780a in my experience) is a lot more nasty. Basically when I tried to Crossfire my 5770s on my Crosshair 2 (Nforce 780a) I could only boot into safe mode. If I tried to boot into normal Windows mode with the cards in and the drivers installed (by installing one and then shutting down, plugging in the other and rebooting) the PC would simply reset as soon as the driver loaded and sniffed the cards. This kind of protection (for Crossfire on an SLI board) is done at a low level in the bios. Meaning that no software can bypass it.
However on AMD based chipsets they are not as nasty. Instead of forcing a hardware reset at a low level the board instead just simply stops the PCIE slots being linked if it finds two Nvidia IDs (hardware dev ids). However, this is where the patch comes in. You plug in two Nvidia cards. You go into Windows and patch it. At which point somewhere deep in your device manager it then appears as if you are using an X58 chipset. Thus, when you load up the Nvidia drivers and it looks at what chipset you are running it comes back with an O.K and will enable.
The way it was done was to basically reverse engineer an X58 bios using a disassembler (no mean feat seeing as all new bioses are encrypted) and finding out just what was happening. At that point they figured it out, borrowed
a portion of that bios and implemented it into their patch.
Due to the way it works it would be *incredibly* hard for Nvidia to stop it working. I mean, what you going to do? stop SLI running on what the drivers see as an X58 chipset?
It's all down to licensing mate. When you want SLI to work on your motherboard you pay Nvidia for the pleasure. When you want a USB port you pay apple $1 in royalties for each one. When you want Firewire you pay apple $3 for each one. ETC, ETC.
Well, that and just being a monopolising Ahole and trying to force people to make a choice. Intel decided to say nuts to the lot of them and just got on with X58. But AMD will likely never EVER pay Nvidia to put SLI on their boards.
Lucid? they don't need to pay any one. Hydra works differently by hijacking the Direct X signals at a low level and combining the processing power of any two or three GPUs you put in but it doesn't link like SLI or Crossfire do. It's proprietory.
So yes. I would say if you're in the slightest bit worried get some cheap Nvidia GPUs capable of SLI (8600gt or something, or, even the old ones like the 7900gt) have a crack and see how it goes. Or, if you already have one 570 then buy another one, if you can't get it to run you can have an 'ahem' accident with the second one and return it for a refund
But yes, in summary, from some one who knows a lot about software coding and programming, I would say it ought to work perfectly