On page 435 underneath the Importance extension, there is a formula for expected ship traffic: S= 10^Ix / H, where H is the expected average tonnage; 100 tons for ports not on a trade route, and 1000 tons for being on a trade route. There is a further modification, I=I+1 for a Busy Empire, and I=I-1 for a Rural Empire. Are these latter modifiers references to importance?
This results in some strange results. The world of Natoko (3209 Spinward Marches B582211-8 Lo) has an Importance of -1; this means an implied tonnage of 10^-1 i.e. one tenth of a ton each week. That's one ship every 1000 weeks at an average of 100 tons; but Natoko is on a trade route, making average cargo capacity 1000 tons, or one ship every 10,000 weeks. That is, one ship every 192 years.
I don't think Natoko is on a trade route, actually. I think that the trade route that it appears to be on is actually a J3 route from Pretoria to Aramis via Teh, bypassing Natoko completely. Natoko may be on a J1 trade route to and from Aramis, but not if the population is 800. And while I think the population should be increased to several thousands to reflect the transients on the naval base, I don't think that even several thousands represent enough of a market to support a regular trade route.
I haven't got a solution, but this is very difficult to square with a Starport B.
Yes indeed. Why (and HOW?) does a population of 800 build spaceboats? I would either change the starport class to D or even E to reflect the amount of starship traffic or change it to C to reflect the Imperial Naval base being nice to civilians and selling them minor repairs and fuel. In no case do I see anyone going to Natoko to have ships or boats built.
We can ignore the formula because it is "Expected" ship traffic, but then why include it in the rules?
The first thing to do, if it isn't already part of the rules, is to highlight one of the most important principles in world-building: "If the rules produce a result that doesn't make sense, change the result." (Followed by the rule: "If you come up with an idea that you like better than what the rules produce, go with your idea."

)
Next, I'd change the formula. There should be a very strong correlation between population and trade. Everything else being equal, a population ten times bigger should have ten times as much trade. Of course, everything else isn't always equal. That's where things like starport class and Importance comes in. A starport class of B should indicate more trade, an Importance of -1 should indicate less trade.
1 Or should it? I'm absolutely convinced that trade correlates with population, but not at all sure how strong the correlation is.
I'm not sure what formula would produce an appropriate base figure, but it should be calculated as Population divided by some factor. How much interstellar trade would 1,000 people generate in a year? If it's 1 dT, the formula should be Population/1000.
Next you adjust for Starport type:
Letter indicates effective starport class
A: +25% trade
B: +10 trade
C: Base trade
D: -10% trade
E: -25% trade
The 'effective' means that if a starport is subsidized by outsiders for some reason, the effective starport class can be less than the actual starport class.
I don't know what range Importance covers, but I think the trade volume modifications should cover the same sort of figures, perhaps -50% to +50%.
Finally make a random roll to account for undisclosed local conditions: 2D-7 * 5%.
Hans