NavList:

You can read a series of messages from Bruce Stark in which he outlines the construction of his tables here in the NavList archive for the Spring of 2003:
Messages written by Bruce Stark, late April through May 2003.

I didn't check to see if he discusses the specific table you're looking at, but there's plenty of detail in there otherwise.

On a general note, while the publication of these tables did a lot to revive interest in lunars among navigators in the late 1990s, they're an aberration, an oddity, an exercise in idiosyncracy, with almost no connection to the long history of lunars, and I would not recommend them to anyone. Stark was creating these tables just before the Internet took off. At that time there were almost no historical resources available to work lunars. His notions about lunars were founded on a prevailing myth that originated in the late 19th century: that lunars were hard because the math was difficult. And the obvious solution to this imaginary problem, this myth, was to re-invent the wheel. It was never the case that lunars were mathematically difficult except with certain poor or overly-refined methods and in certain licensing exam circumstances where difficult cases were emphasized. The world did not need yet another exotic, mathematically opaque method for clearing lunars in the 1990s. 

If you would like to learn a good method for working lunars that is simultaneously, easy, fast, and accurate, look no farther than Thomson's tables, which were published by their creator, David Thomson, in many editions and were also adopted (or, less charitably, "stolen") by Nathaniel Bowditch and added to the New American Practical Navigator beginning in 1837. You can find links to these works and many more through an index maintained on the NavList website: Bowditch and More: Historical Navigation Books Online.

There are other tools for lunars on my website: http://www.reednavigation.com/lunars/.

Frank Reed
Clockwork Mapping / ReedNavigation.com
Conanicut Island USA