Well partly I'm totally out of touch, this thread has refreshed my understanding a bit. I was probably focusing more on physics than other sciences. I was actually looking up PhD thesis from friends written back in 2002 and asking myself if any of this research was actually ever used.
It was also partly about funding and why research should be funded for research that's not ever used.
.
Because whilst it may not be used
now it can and often does lead to uses later on, or a better understanding of how things work together.
You don't just research the "useful" stuff because until you've done the research you don't know what is useful, or your research may not in itself be useful, but may increase the knowledge of a subject - for example who really needs to know how a dinosaur lived, we're never going to have to feed and care for them, except that by researching them you find out about the conditions of the planet at the time, and how animals evolve.
Research into space for example was largely "not useful" by industrial reasoning at the start (it was largely two superpowers having a contest to show who was "better"), right up until materials were found due to the needs of space that suddenly changed how we did things like cook, or build cars, or communicate, or to give an idea of how research in one area for commercial use opens up another completely unrelated one, IIRC the abundance of cheap micro drones owes a lot to the mobile phone industry where for the first time there was a real push for ultra miniaturised high performance motors that used small amounts of power, and for them to be made by their billions - a lot of the motors in drones can IIRC trace their ancestry back to the motors used to make your mobile vibrate.
My friend who works as a researcher has done stuff that has ended up being industrially viable (but not yet implemented AFAIK because it needs other stuff to catch up), he's also worked on stuff that turns out not to be viable yet but in 5-10 years when equipment that can make it is not just confined to a lab... and stuff that may not ever see the light of day outside the lab but might also end up getting used next week if someone somewhere needs to do something and it's the solution.
A heck of a lot of "pure" research is about finding out if something is possible, then often another research team will work with that to make it into something useful.
Even a lot of "silly" sounding research that gets done often has a reason, even if it's just understanding something like how the brain reacts to things and in what order (a study into laughter for example sounds silly, until you realise it helps map the brain function out).
Basically you never know when something you're researching is going to be useful, which is why you fund speculative and "pure" research as well as the obviously commercially viable things - things like particle accelrators and cold fusion research may not seem to be getting anywhere fast (in part because it's so costly and takes so long), but you can be fairly sure that along the way they've been making advances that quietly help in other fields, if just because the tools they need may not exist yet, nor not be accurate enough.