With the caveat that nobody can really predict the future with certainty, I'll take a stab at it anyway:
-Prices on LED diodes will continue to go down, with per-diode performance going up. But since diode cost only represents a relatively small fraction of the total cost of a panel, by itself this is probably NOT going to translate into a proportionate decrease in panel cost. This is particularly true since the market for grow-specific lights is pretty small.
-With a *lot* more cumulative growing experience and evolution of design out there, manufacturers of panels are going to be putting out better and more refined panels. Combine with cheaper LED diodes, and we're talking panels with more optimized spectra, and with more efficient energy usage.
With better designs out there the quality of the copycat-knockoffs will also probably improve too over time.
Add these things together, and I think the net effect will be that over time panel performance will continue to improve, and prices will continue to go down, but I don't think we're going to see and sudden and dramatic improvements in either performance or pricing in the next few years. There will be slow improvement in both.
Like any new technology, the better it gets, the more people will use it, and the more people that use it, the cheaper it gets. All that said, I still don't see LED technology entirely replacing HID lighting for growing anytime in the next five years.
HID technology is mature, and its use is widespread for all sorts of non-gardening applications. The lamps and bulbs are relatively cheap, easy to maintain, available everywhere, easy to service, and highly effective for growing. Even though HID lamps use more power and generate more heat than LEDs, those are really the only two disadvantages they have at this time. LEDs make sense right now in environments where heat is a big concern, but they are still going to have to improve by quite a bit in terms of cost before they entirely displace HIDs.
On plasma lighting, that's a chicken-and-egg thing. I think the performance of plasma already beats HPS, but the technology is still so new that costs of the plasma units themselves aren't competitive.
Remember, there is a huge functional base of HID lights in active use right now. At current HID lamp costs and with current energy costs it doesn't make sense for people using HPS to make the switch now. But as more and more people use plasma lighting, because of economy of sale the per-unit costs will go down, eventually becoming competitive with HID lighting in many applications.
Put more simply, once it becomes cost-effective for people to start putting plasma into their warehouses, and lighting their buildings with it instead of HID lighting, then it will be equally cost effective to use it for growing.
With plasma or LED, the cost of energy is a major factor too. If energy costs go up, then the relative competitive advantages of LED and plasma increase proportionately. In places with insanely high energy costs (ie powering the space shuttle, for example!), then these things are probably cost effective right now.
On lasers. . .forget it. Lasers are highly energy efficient, but they're also relatively fragile, bulky, and expensive compared to conventional LED diodes. I don't see any commercially viable laserlight growing systems hitting the market in the next 10 years.