Firstly, these tables are useful for more than just class recommendations. They congregate the more refined advice throughout the book into a simple shorthand that makes for quick decisionmaking, and can quickly congregate potentially multiple pages of paragraph of not-always-needed text into a simple table segment no more than a couple of inches squared. I don't need recommendations of mechanical effectiveness as much as I need a measure of "Is the option recommended?: Very likely/likely/not likely"
Secondly, Starfinder will grow with time, Even if Pathfinder is not included, it'll eventually have options over the years. I don't think it's a stretch of an assumption that Paizo will not stop Starfinder at 8, or even 12. Always room for more, plus class books sell. Now whether Starfinder will be a 1 class a year or a 2 class a year game once it catches up to 1E's roster, time will tell.
Thirdly, Starfinder is compatible with Pathfinder, and includes Pathfinder content. Even Guilt of the Grave World asserts this as two of its 4 recommended ancestries are the dwarf and the orc, which will only be available in Pathfinder for quite some more months. Starfinder is proving itself willing to use Pathfinder as an expansion pack, and therefore, can make recommendations from Pathfinder when convenient, so Starfinder already is not a small game, and recommendations can be made around the mechanical effectiveness of specific Pathfinder options, as well, if a low Starfinder option list is an issue. For example, while Pathfinder classes might not need detailed paragraphs to explain their usefulness, they can still be included by placing particularly useful ones or unuseful ones under the table as a shorthand. In real play, Pathfinder stuff gets used. I already have one player who plans to be a mech pilot by way of being an Inventor, for example.
