Additional Question: What do you think of "If it ain't broke, don't fix it"?
I want to replace some of our headlights and light bulbs with better ones.
It is more of a "What is the gain" vs "What is the cost". The gain of replacing something close the end of it's life could be no need to mess with it in a hurry when the old one fails, doing it's task better,... The cost is the remaining service life, but as well in case of really early upgrade the wear of all the attachment screws and latches and the consequent risk of breaking something by some mistake during the work or so.
The questions should be: Is it doing it's job? Is the known state so, it is expected to do it's job safely (it won't threaten anything else by it's expected failure) and reliably enough still for reasonable time (so it won't need any extra expensive service-man "mission"; that its mainly related to installation where you need to rent some special equipment, needs more time to get there and so on)?
If the answer for both is "Yes", then there is usually no reason to mess with it at all.