The title pretty much says it. We know fermenter temperatures can be well above ambient, but which range are the yeast manufacturers referring to?
Jumping off from some of the comments already posted:
I think to make sense of this it helps to realize that, for those brewers who produce most of the beer in the world and who are probably the most significant portion of these yeast manufacturers' business (that is, commercial brewers), the ambient temperature is, for the most part, basically irrelevant.
As an example, consider the fact that many big breweries ferment in massive outdoor tanks. If you left what was going on inside (fermentation) open to the influence of the ambient conditions it would be nearly impossible to create beer consistently (and breweries don't get as large as you have to be to use such tanks by making inconsistent beers). So the tanks are obviously insulated from the outside conditions. But then of course, you need a way to remove the substantial waste heat of fermentation which would otherwise now be trapped by the same insulation, so the tanks (and their contents) are cooled mechanically. You can kind of see then, that in this situation, the measurement of interest is necessarily going to be the temperature of the fermenting beer inside the tank, not that of the conditions outside the tank, conditions which (should) have very little, if any influence on the process going on within.
Of course on the home scale, while many strive to match the consistent quality of commercial breweries, we are limited, by cost, space or what have you, in what equipment we have. But, regardless of what we use, the quality of the resulting beer is determined by the particular variables that any given batch experiences while in process. That is, the beer produced will be influenced by the temperature because yeast (which behave quite differently at different temperatures) experiences the same conditions as the liquid in which it's suspended. Certainly the ambient temperature will, in the absence of insulation, cooling, temperature measurement and control devices have a dramatic effect on said liquid temperature, but this ambient temperature is not a process variable in itself, it merely influences another variable (fermentation temperature). If you are limited to only being able to control ambient conditions (and not actual fermentation temperature) you really do have to accept that you're not actually controlling a variable relevant to the outcome, you're controlling a variable which influences a variable that is relevant to the outcome.
I understand why you are eager for a cited reference regarding your question (and perhaps somewhere out there an authority might actually make the distinction) but it all seems much more basic and fundamental than that to me. Think of it this way: if you're making a roast and you want it to be medium-rare, you don't stick the thermometer in the oven to find out if it's done, you stick it in the roast.
They are referring to the temperature of the yeast in the fermentor.
Setups and equipment vary so much that the ambient temperature isn't a good indicator. Brewers often cool or heat a fermentation vessel to achieve the right temperature inside for the active yeast.
Cooling solutions range from a cooler basement, to standing water plus a fan ("swamp cooler") to very expensive high-tech products. Heating solutions often involve some sort of electric heating pad.
Whether heating or cooling a fermentor, good insulation is helpful because rapid temperature swings are bad for yeast in addition to out of the ideal range temperatures.