Sure Optical is faster now but if history shows us anything they will be able to get copper up to that speed in a short time.
Copper will not catch up. The only electrical technology that is known about that has a hope of beating optical is carbon nanotubes, and that is a long way off. A single copper cable simply cannot contain such high frequency signals over any useful distance without a huge input of power, which will most likely lead to the copper melting. Optical systems do not have this problem.
There's no point sending the data through optical just to have to convert it back into electrical charges.
Not true, the shorter the cable, the less the inductance, and the less power needed to transfer super-high speed data. That means wires traveling between two CPUs (eg. QPI) can transfer more bits/s than similar wires connecting the CPU to your external SSD can without melting. Very short distances can remain copper, and will for a while at least.
It's 3 bites at the sales cherry instead of one.
Good point.
The question is how long will it take for this technology to reach the internals of a system, allowing companies like alienware to 'play' with the light in expensive, eye-candy gaming rigs.
Reaching the internals of a system, I wouldn't expect it to be long. Computer companies can have a bundle of copper wires leading from the CPU to the GPU, or just one or two Light Peak interconnects. Miniaturise Light Peak further, and you can print everything right on the circuit board, along with optical waveguides right there on the silicon, no more cables. I've worked for a research company successfully doing this, but the progress is slow. So slow you wonder if other technologies, like nanotubes, will beat optical technologies to the on-chip arena.
Oh, and the wavelengths used to transmit data are all infra-red. They are for telecommunications too, and that is unlikely to change due to efficiency and optical attenuation. So sorry, but no light shows.