The mutual exchange is a structural relation rather than a positional one, and that gives the class a distinctive property: it does not depend on the two planets being near each other, aspecting each other or occupying any particular kind of house. Two lords sitting in each other's signs anywhere in a chart form the exchange, which makes the class easy to identify and hard to overlook.
The classical grading into maha, kahala and dainya turns entirely on which houses are involved, so the same geometric relation is read as favourable, mixed or difficult according to whether the sixth, eighth or twelfth is one of the pair. That is a clean example of how the tradition assigns value to house rather than to configuration, and it is why an exchange cannot be judged from the fact of the exchange alone.
Timing applies as in every other class. An exchange is expected to show during the periods of the two planets that form it, and because an exchange binds two houses together, the tradition reads the result as affecting both at once. The atlas records the structure and the classical grading, and leaves the reading of a particular chart to those trained in it.
There is a further reason the class is prominent in practice. An exchange is one of the few classical combinations that can be identified from a chart at a glance, with no calculation of strength, aspect or distance, so it survives the transition from manuscript to modern software and from specialist to general use better than most. Ease of identification is not the same as importance, and the classical grading exists precisely to stop the two being confused.