The significance of minifigure-scale varies dramatically between subjects. Larger starfighters such as the T-65 X-wing or TIE/sk Striker have achieved near-perfection and that is apparent simply when looking at them, without needing any calculations. Other vehicles deviate slightly from the source material, but are ideally-suited to interaction with minifigures.

Using that figure and the following simple calculation, we can ascertain the approximate minifigure-scale for every vehicle. I have selected ten sets which I believe to be among the worst ever released and have tried to explain my reasoning for each selection below. The following items have not been taken into consideration as I do not think they can be properly compared with conventional System sets. That’s the awkward thing about minifig scale. The Earhart plane has a fuselage that’s just four studs wide, but the bare minimum to seat two minifigs side-by-side is six (and even then they’ll need to roll down the windows so they can rest their arms on the sills). Based on height, cars scaled to a minifig end up looking like the polybag Elva that can’t even fit a whole minifig. Scale the model to their width, and that’s how you end up with Speed Champions producing 8-wide tanks.

Notes: Canon sources sometimes give a length of 23.40 metres, but that describes the fully armoured BTL-B Y-wing Starfighter deployed during the Clone Wars, rather than its later stripped-down equivalent. However, I think it would be unfair to ignore sets from the first few years of LEGO Star Wars entirely. Sets should be judged according to the context of their release and in that regard I think certain older models, such as those mentioned in the article, fare comparatively poorly in relation to other sets released in the same year. Also, from a purely practical perspective it would be almost impossible to identify a single year after which we are able to judge sets by modern standards. This is absolutely spot on. The perception of minifigure-scale is paramount, regardless of actual accurate scale and particularly when the characters are exposed inside the vehicle. Measurements taken from StarWars.com, Wookieepedia.com or the most recent canonical reference book.

As for "accountants had taken over Lego and people only care about making money instead of bringing joy and creativity to children", any company nowadays is about money. Like it or not that's the world we live in. LEGO can't make sets if they don't bring in the money to do so. Star Wars brings in the money, and quite a lot of it.

As for the Flagg, at 6’ long, that would only be 115’ long. Even helicopter carriers seem to start at about five times that length. The USS Gerald R Ford is right around 100x that length. I can’t find any hard numbers, but based on photos, I think I did find one “aircraft carrier” that was shorter. In 1849, the SMS Vulcano became the first “balloon carrier”, when it was used to launch bombs carried by hot air balloons against Venice. Over 40 years earlier, the 127’ HMS Pallas was used to launch kites bearing propaganda against Napoleon. So, when you get down to the length of a football field, you’re largely talking about unmanned aircraft, or at max maybe one small helicopter. Once again, this issue can be attributed to the limited selection of colours available. Dark green pieces are ideally suited to Boba Fett's Slave I and have been used on every subsequent model of the vessel. Unfortunately, they did not appear until 2003 so the designer had to use the standard shade of green which is not at all suitable. This model also suffers from a severe scaling issue and its flaws become even more apparent when the set is compared with 7153 Jango Fett's Slave I which was released just two years later. 7257 Ultimate Lightsaber Duel

I definitely fall into the camp that believes comparing the builds from the first few years to sets from several years later is something of a pointless exercise. As already mentioned numerous times, the color palette, available parts, and experience in creating these sets was at a much lower level than later. For the most part, the best you could hope for was an approximation of what was being created. I think each of the three sets listed here actually do a good job in that respect.

At least in Lego form it's a lot easier to mask scale inaccuracies, because there is no actual solid point of reference scale wise. A minifig is intrinsically a really bad estimate of the human form factor, so everything else is a lot easier to sell as "somewhat in scale".