for all the things you do...
Color really is a "problem" for the collector of classic stamps. In many cases collectors are trying to identify a color based on a written description in a catalog, a task that is clearly impossible. In the absence of a good point of reference, how could one distinguish between carmine, red, pink, rose, rose red, or carmine rose (actual choices to be made when identifying variations of the "2 Cents 2" red George Washington issues)?
In other cases the collector might have some sort of printed color chart, which is a quantum improvement over written descriptions, but flawed nonetheless. Even modern printing methods are not consistent enough to accurately reproduce the colors and shades in question - at least not economically. So while the color charts can point you in the right direction, there are many times when no color on the chart is an exact match to the stamp being identified. At very advanced levels, charts using colored tiles are available, however these references are often for a very specific stamp issue, great for the specialist, but still leaves much of the problem unresolved.
To better understand the color problem, we need to look at its roots. Back when the world was young, and stamps were either single colored or bi-colored, pioneering philatelists not only cataloged major color variations - such as a change from blue to green - but minor variations or "shades" as well. Herein lies the rub. The differences in these shades are often extremely subtle. They are not the result of errors or intentional change of color, but rather normal variations in the ink that was being used. Sometimes a different manufacturer's product was employed, sometimes the ink in the press was running low, and sometimes the ink color was simply inconsistent from lot to lot. Take any 19th or early 20th century stamp issue and line up 100 copies and you will see that there are not simply two or five color shades but actually a continuum of shades from pale to dark. This is particularly true of stamps printed with red ink, which is notoriously unstable.
To complicate matters further the colors of stamp inks have been known to change over the years. Fading of colors, especially if the stamps have been left sitting in sunlight, can be extreme. Purples and violets can easily fade to gray. Sometimes chemical changes in the stamp will dramatically alter the color. These stamps are called "changelings." Yellow and orange inks are very prone to oxidation and will change to brown.
So what is the well-meaning collector to do? I've heard more than one person say that they choose to simply ignore the more subtle shades. While this is certainly a reasonable option, I don't think it is a satisfying one for most. To the extent that philatelics is a game of sorts, I personally hate to leave any known variations out of the equation. If the catalog says there are two varieties to collect I want to find and identify them both!
Bearing in mind that the color problem is intractable, there are some pointers that can be helpful.
As in so many cases, one of the most helpful things you can do for yourself is to build your collection in a well-ordered fashion and use your own stamps for reference. Nothing demonstrates the difference between "carmine" and "lake" better than setting a US #219d next to a US #220. In general, stamps from the same country and same era are likely to have consistent catalog descriptions - i.e. the yellow-green of one series should be similar to the yellow-green of another series.
When identifying tricky stamps, make sure that you have ruled out the easier distinctions. Have you measured the perforations? Have you checked for a watermark?
The lighting of your workspace can make a tremendous difference. Obviously your workspace should be well lit, but also the type of lighting is important. Natural light (i.e. sunlight) is full spectrum and consistent, however natural light is often unavailable. Avoid, if possible, fluorescent lights because they can really make shades difficult to distinguish. I use a good old-fashioned gooseneck desk lamp with a 60 watt bulb. There is a brand of light on the market called an Ott-Lite lamp which purports to reproduce the natural light spectrum. Many people swear by this technology, not just in stamp collecting but in many other fields where accurate color distinction is important.