There is no single answer to that question: it is going to vary program by program. Some shows, like America's Test Kitchen, are going to be quite precise--at at least, they will intend to be. Other shows, not naming any names, less so.
The thing is, with very very few exceptions, you should not be cooking to time anyway. The time in a recipe is just a guideline so you have some expectation for logistical planning, and to know when to start testing.
Quality recipes--and quality television shows giving you recipes--should always give a test for when a food is done. This can vary from "pulls away from the side of the pan" for brownies, to "pulls apart easily with a fork" for a braised pork shoulder, to "internal temperature of 135 F" for a roast.
Even for shows that intend to give you very good time estimates, testing is still important, because there are many variables that effect how long it takes food to cook. For something as simple as a steak, the starting temperature of the steak, the thickness of the steak, the starting temperature of the pan and its thermal mass, and the power and setting of your burner (hob) will all effect how long it takes the steak to finish. You probably don't have the same room temperature as the test kitchen where the recipe was developed, the same stove, the exact same settings, and the same thickness of steak.
Such variables apply to almost all recipes. The test is what tells you when to stop cooking; the time is only an estimate.
As an aside, the few things that must be cooked to time are those that are 1) uniform in size and cooking properties; and 2) foods for which it is difficult or impossible to observe the effect of the cooking directly. The only two examples I can think of that are appropriate are cooking eggs in the shell, and roasted peanuts in the shell.