Currently, Cadence ArrayValue, CompositeValue, and DictionaryValue
creates atree.Storable directly from StorageID of its Atree values.

In the future, some Atree values can be inlined so they don't have
StorageID, and Atree values need to handle Storable depending
on maxInlineSize and other factors.

This commit delegates Cadence composite values' Storable() to its
internal Atree values.  So future changes to Atree values and their
storables don't break Cadence.

Currently, storage address is derived from StorageID().

In the future, some Atree values can be inlined so they don't have
StorageID and address needs be derived differently.

This commit delegates storage address to its internal Atree values,
by calling Address().

Currently, Array.StorageID() and OrderedMap.StorageID() are used as
identifier to track resources, etc because storage IDs are guaranteed
to unique. However, atree storage ID should be only used to retrieve
slabs (registers) from storage.

Also, when Atree register inlining is implemented in the future, some
resources may not be stored in separate slabs, so they will not have
storage IDs anymore.

This commit uses Array.ID() and OrderedMap.ID() to uniquely identify
resources.

…type-test

Fix data race in sema/type_test.go

…site-storable-to-atree

Delegate Cadence composite `Storable()` to Atree

…atree-storage-id

Delegate storage address to Atree