Some people talked about tweaking the world generation system, and there's no doubt that there are things that could be done to improve it. Minimum world sizes for various atmospheres and hydrospheres, minimum poplulations for worlds with shipyards, minimum tech levels for worlds without breathable atmospheres, etc.
But fixing the WGS, worthwhile though it may be[*], will not solve the existing problem with the canonical OTU UWPs. The problem with them is that they were published without anyone vetting them first.
[*] And I'm not entirely sure it is a good idea, since it would eliminate strange combinations entirely. The problem I have with the canonical UWPs of the OTU isn't that they include some impossible and near-impossible UWPs; it's that they include so
many impossible and near-impossible UWPs (as well as some that are perfectly possible, but unreasonable in the context they exist in).
Don't get me wrong. I like the world generation system, I really do. I think it represents a brilliant idea. But what it produces is not a finished product.
The Spinward Marches was not a finished product. Even if every single UWP in the Marches[*] had been individually possible, the odds are that the game universe could have benefitted immensely from having some of them altered a bit. Because the Marches are not 400 individual star systems jumbled together randomly[**]; they're 400 star systems that has interacted for a thousand years and more.
[*] I use the Marches as an example because that's the part of the OTU that I'm most familiar with, but everything I'm saying will apply to any other published Traveller game setting.
[**] Well, yes, they are, but they shouldn't be

.
Take a perfectly possible UWP like C866546-8/1 (the '/#' is the way I denote the population multiplier when I don't care about gas giants and planetoid belts). That would be perfect possible in a setting where there are lots of empty worlds just as good available for other settlers to pick or a setting where there is a powerful interstellar state around to prevent would-be claim-jumpers from moving in. But it's pretty damn unlikely in a setting with a history of centuries worth of immigration and no nearby worlds with the strength to defend it (especially if there are few other nice worlds in the neighborhood). Or a crappy world with a high population could be explained as a trade centre, but only if it actually happened to lie between the proper distance from two high-population worlds.
IMO,ideally no official set of UWPs should be published without someone having gone over every single one of them and found an explanation for each.
Now, I know as well as anyone that this simply isn't realistic for a game company that hopes to make money. I know how much work was involved in making the 30-odd worlds of the Sword Worlds subsector work together as a group. Sure, some of the trouble we had stemmed from impossible UWPs, but my guess is that 90% of the work came from making perfectly possible UWPs work
together. I've also, as a hobby, been working for a decade on the history of the Spinward Marches. My ambition is to make a Galactic map for (roughly) every century from 55 to 1120, and so far I've managed to get the Year 55 map fairly well finished, the Year 125 map more than halfway finished, and scratched the surface of the Year 200 map.
So, no, I don't expect any game company to pay anyone for doing the Imperium in that level of detail. But I do know that a 'first pass' of a subsector can be done in about 8 hours (because as an experiment I did the two subsectors in
The Stoner Express in two days. Never got around to posting it...).
Of course, the entire Imperium has about 300 subsectors, so I don't know if even that much is realistic.
Hans