I believe they are referring to the cumulative elevation gain [0], which is essentially the sum of all increases in altitude during the route, without subtracting loss of altitude by descent. While in theory this implies potentially different values depending on which direction the route is taken, overall the difference between both values is just the difference in sea level between each end of the route, which may be negligible.
No. They are useful as a representation of the overall effort one has to make. The effort made in ascending 500 metres is not recovered when you descend 500 metres. Ascending 1000 m continuously along a 10 km trek is significantly less effort than ascending and descending five 500 m hills along the same distance, even if in the first case your end point is 1 more km above sea-level than where you started, while you remained at the same altitude in the second case.
? Walking up the stairs in a 1,000 ft tall skyscraper is 1,000 ft of elevation gain. Doing it twice is 2,000 ft gain. Certainly it is useful to know whether the gain is 0, 1,000, or 2,000 because those will call for different amounts of effort and physical fitness. This is the same type of gain expressed in these figures.
