Used for Webdriver 3 multi-touch gestures
See the Webriver 3 spec https://dvcs.w3.org/hg/webdriver/raw-file/default/webdriver-spec.html
The MultiTouchAction object is a collection of TouchAction objects
(remember that TouchAction objects are in turn, a chain of individual actions)
Add multiple TouchAction objects using the add() method.
When perform() method is called, all actions are sent to the performsTouchActions.
The performsTouchActions performs the first step of each TouchAction object simultaneously as a multi-touch
"execution group". Conceptually, the number of TouchAction objects added to the MultiTouchAction
is equal to the number of "fingers" or other appendages or tools touching the screen at the
same time as part of this multi-gesture. Then the performsTouchActions performs the second step of each
TouchAction object and another "execution group", and the third, and so on.
Using a waitAction() action within a TouchAction takes up one of the slots in an
"execution group", so these can be used to sync up complex actions.
Calling perform() sends the action command to the Mobile Driver. Otherwise, more and
more actions can be chained.