These are difficult questions to answer.
First, the product "LED headlights" is not specified for a Y2k anything, since they did not exist at that time. Therefore, any such product that might be considered will have at least some aspects that can cause trouble. For example, would such a product be a large number of low powered devices, or would it be a few high output devices as recently made available? Would they fit properly in the available installation without adverse light leakage? Are they sufficiently heat-sinked to provide a decent longevity of the product? Can the devices (individual LEDs) actually do what they are advertised as being able to do? Will the power demands make it possible to drop in and not confuse the relays and ECU (central convenience module?) of the current car?
Nevertheless, if all aspects of the power sourcing for headlights are known, as well as what the power demands of the actual light source devices are, then perhaps a good decision can be made.
One of the big problems of LED lighting is the initial cost. Small LEDs are cheap, but they are made to be small indicator type lights and don't put out much light at all. Heat is not a big deal, because any time you are using a device where the power consumption is measured in just a few milliwatts, you can probably get away with counting on rejecting that much heat through the wire leads. When you step up to the light levels required to usefully illuminate something else at night, then the power consumed by the device will go up in 2 or more levels of magnitude. Now you are talking watts, not milliwatts, and heat rejection becomes a serious problem. These are not incandescent bulbs which are made for high heat. These are semiconductor devices that are killed by heat levels that won't make an incandescent bulb glow a dull useless red.
So, high wattage requires high heat rejection. Better make sure the LED lights have a serious ability to reject reasonable amounts of heat.
And beware the second killer of aftermarket LED devices: that of using crummy devices to begin with.
Seriously, there's been zillions of LED devices produced for such purposes as you desire that are just plain POS. Cheap, they obtain brightness by overdriving a small device. The package will advertise tens of thousands hours of life, when reality is tens of minutes until they fail miserably. Most expensive junk you'll buy this week.
My advice is to find good devices, pay the money, and do some thinking so that you'll be able to count on them lasting more than a few weeks rapidly diminishing brightness before sadly premature failure.
I decided to make some of my own for my 4-dr when the city lights (that small bulb in the bottom of the high-beam reflector) blew and were stupid-difficult to replace. I found some lumileds, I soldered 3 amber devices (tried some other colors as a fun experiment, but really the only considerations were white or amber) to a copper tube as a heat sink, rigged a power supply inside the headlight enclosure, and sealed it all back together. In the time since 2006 when I did it, no failures and the brightness remains as new. You can see photos in my personal album if you care to look.
Anyway, my point is that you need to study the device(s) in question. Usually, for the year you are thinking about, then answer is probably no, don't do it. You'll not be pleased with the long term results. I could be wrong, of course. I tend to be cynical and pessimistic after having seen too much C.R.A.P. floating around that wastes your money.
Cheers,
PH