They bundle a copper wire along as well.
No they don't. The initial implementations and plan was not to include power. Putting in copper and getting into the power distribution business inhibits one of the features they were going for; distance. Stuff like USB , PCI-e , display busses and to a very large extent even FW (although I doubt they are trying to address FW with Lightpeak) only work over distances like 10m (or less). Lightpeak is suppose to work around 100m.
Additionally, wrapping two disparate cables into one (a fiber and a copper one) will tend to make the cable more expensive and at least thicker.
Powering other devices is a short haul problem. That's at odds with want they wanted to solve. Over very short distances the native protocols work fine. You have a disk sitting inside a computer... you'll use SATA. You won't convert to LP then back to SATA just to run it 6 inches. Likewise if are plugging a USB device directly into a socket on the box.... why add extra costs to convert USB signals to/from LP? All that does is just add costs. No speed improve.... it just costs more. Who wants that? Most people don't want something that is gratuitously more expensive.
If the disk is 5 feet and the USB device is 30 feet away... yeah sure. LP have huge leverage and benefits. That is one problem space Intel aimed this technology at. Likewise if your monitor with its USB hub and cameras/mics are feet from the computer. One wire multiplexing several protocols into one data stream.
Intel said they'd look at maybe putting in copper. That certainly helps with building the "one ring to rule them all" hype that people want to spread about LightPeak. It is a cheesy experiment to do for them since they can just pull the USB 3.0 fiber experiments off the shelf. (just like pulled elements of that off the shelve to do Lightpeak itself. ) USB 3.0 using fiber for the superspeed bus got rejected by a significant number of device implementers due to higher costs. I don't see what about lightpeak makes that better. The laser costs are down but they and cable costs are still higher.
If the objective is to be, as stated by Intel, to be complementary to USB then copper isn't necessary. USB sockets can solve short haul problems and LP the long haul and greater than 5Gbps ones. Makes no sense to try to build one system that spans the entire continuum from mouse to 10TB disk array box. That's a "hammer meet nail" mentality, that is grossly myopic.
What you tend to end up with is a system that does less of everything, than be better than most in a smaller solution space.
still refuses to adopt eSATA. It is not happening and I am not complaining. Do not like it, buy a PC. 
