In a nutshell, think of low data as what the compiler produces based
on the information declared in your state files.
To be somewhat more detailed:
* Each piece of information in an SLS file is a piece of high data
* SLS files are combined together to form a high state
* The compiler converts a high state (how the user thinks of it) to a
low state (how Salt thinks of it)
* A low state is divided into low chunks, which are ordered by
dependency, then user-declared order, then lexigraphically
* A low chunk represents a call to a single state module
* The state module then uses that data to perform a call (or calls) to
an execution module
This is pretty low-level stuff... but the low data is close enough to
what we see as users, that it should be usable to many less technical
users to understand what Salt will be doing in a highstate call, and
in what order.