John,

The use of a view should rightly be reserved for their proper use, which is generally to take a complex table relationship and allow it to be used like a table, and/or present a limited subset of things to help protect a set of base tables, as part of either a reporting set up, or as a participating method in an overall security mechanism. The use of CTEs is generally a good idea when separate, complex elements must be JOINed. A CTE is, effectively, an in-line view. The optimizer often benefits from this kind of query construction because it can look across the final query's JOINs and determine a better query plan by limiting each CTE element based on those JOINs, whereas a sequence of sub-queries rarely benefits from this kind of optimization.

EDIT: No element of the original query was sufficiently complex to justify the use of a CTE, as even the sub-queries weren't necessary.