OP eating hamsters.
Plastics - Lots of time and money went into studying plastics and research about their molecular structure before they became a really huge industry.
Steam Power- Took decades of experimentation and research before really viable engines came to market in a big way for a variety of uses.
Airplanes - Took decades to get to the manned flyer stage
There are lots of other examples plus you have to account for convergences. As technologies matured they allowed advances in other technologies. The bubbling critical mass of existing technologies in modernity allow other emerging technologies to come to market and have applications much sooner than they would have in the past. Carbon nanotubes are a perfect example of this. From discovery to real word applications in less than 20 years.
Lasers: Einstein first proposed stimulated emissions in 1916. First optical laser built in 1960. A lot of work in between (including Masers in the '50s). It still took several years for really useful things to be done with lasers after that.
Everybody and his SciFi writing uncle had ideas about artificial satellites for centuries. They were Goddard’s goal from the early 1900’s. Serious work was done by Germany during WWII. Wasn’t until 1957 they became a reality.
And I suppose you could include advances in medicine, such as cures for cancer, that were pursued for decades before coming to fruition.
Photocopy machine. Developed by one persistent engineer who had a lot of problems to overcome. I believe he worked for a company named Haloid, which changed it’s name to Xerox.
Flat-screen TV. Finally practical for the real-world market after decades of hope and hype.
Computer voice recognition has finally become a useful tool instead of a gimmick.
Garr!
Sigh.
Anyway, you guys are well into the thick of answering the OP. Essentially, I was concerned with “pie in the sky” research that promised big rewards if it worked, but required decades of preliminary work with almost no practical returns at all.
i.e.: fusion, nanotubes, high-energy beam weapons, high-gain photovoltaics for large-scale energy production, high-temperature superconductors, etc.
How long should scientists work before they give up? After all, the A-bomb didn’t even take that long (and yes, the massive govt. investment in the Manhattan project is duly noted).
The main counterexample I could think of was powered flight.
I would tend to rule out anything that provides incremental success–like refinements in combustion engines, improvements in rockets, all computers … even fuel cells to an extent, since NASA depends so much on them. Though commercial fuel cells (i.e., for cars) are still pie-in-the-sky, despite GM’s current billion-dollar investment.
Yeah, those OP eating hamsters are a really advanced technology. 
Nuclear power was another one that took decades to move from discovery of the basic principles to actual use. If you count discovery of Uranium (1789) as the beginning of atomic research, it was 156 years of research before first utilization (1945) took place.
It finally paid off, so I’m confident that fusion will too, once they figure out a way to contain the reaction.
Maybe I’m defining research a little too broadly. You could also say research into nuclear power & weaponry started at the discovery of the alpha & beta ray (1898), the gamma ray (1900), radioactive decay (1903), or the first accurate model of the atomic structure (1911).
Even the latest starting point still leaves 34 years before the first atomic bomb was detonated. The atomic bomb wasn’t just 2 years of study then bang! a bomb! There was heavy research going on for a really long time beforehand that led up to it.
I would put the starting point as the origination of the idea that a bomb could be made:
http://nuketesting.enviroweb.org/hew/Usa/Med/Discfiss.html
Conception to operation, 13 years; half of that as increasingly active research.
Application for nanotubes: X Rays to Go: Carbon nanotubes could shrink machines
I forget exactly when they were discovered, but wasn’t it about 15 years ago?
The lightbulb?
PLease elaborate. How long did we pursue that one?
Now that’s cool. Good ROI.
Somewhere around 1970 (sorry, no cite) some scientific commission called for the perfection of a navigational aid which worked independently of the earth’s magnetic field and/or astronomical observations.
Today, my cell phone has a GPS chip in it. GPS wasn’t “perfected” until about the mid-1980s, as best I can tell, so call it fifteen years. I think it’s a pretty good example because it was a very pie-in-the-sky idea with minimal practical value until the complex system was put into place and de-bugged.
http://www.ushistory.net/toc/electricity.html
Which lead the way for the creation of an electrical power system, and thus major improvements in dozens of electrical apparatuses, especially the dynamo.
http://outreach.missouri.edu/ceupdate/scripts/1997/10/defeat.htm
Even if you use Edison’s practical bulb (1880), you get 71 years of work by many people. Including through Coolidge gives 101 years.
I thought a woman invented the photocopier.