It depends on the programming of the "app" (or .exe or .dll) that accesses the registry. Some will create keys/values if they are not present.
But in this case, as well as with many Windows "tweaks", the default is meant to be not present, instead you alter the value by creating the registry tree above it and then the value.
Most often these "tweaks" came into existence because someone somewhere asked for it (Microsoft's Volume license customers? or some team) and is not meant to be widely known to the "unitiated" mostly because it adds another branch to eventual support efforts (and support costs manpower and money), but most importantly, because it keeps them in control.
That's why you see in the forums, very often, the similar question: "The registry key doesn't exist. What should I do?" Well, you create it, of course.
Registry reads also take two forms, also depending on the programming (the function they call). 1. they stop at the highest branch if nonpresent 2. they ask to read the full path of the key/value directly. In case of (1.) you need to create registry branches one by one until you arrive to the final value that is requested.