What is the difference in interpretation between using one house system in preference to another, other than a possible change of planet, sign or house?
No difference, usually!
Outside of different approaches to astrology (like Vedic vs. Western) houses generally have the exact same interpretation, which means changing the house system doesn't change the interpretation (outside of planets changing house, of course). Anyway, there are two primary reasons why one tends to consider a house system to be "good":
(1) Experience: One may believe a system is good based on what other presumably reliable sources have said, or on their own personal experience. This can be subject to error, for example believing that something is true just because others think it's true. (Consider how the "best minds" believed the world was flat for a long time, and that the Sun orbited the Earth, until it took revolutionary thinkers to suggest otherwise.) Similarly, one may have had interpretation results that seem right for them, but they're based on a small sample size, so others who have worked with a different (also small) set of charts may have been drawn to a different system. Until detailed statistically relevant studies similar to
Gauquelin's "Mars Effect" are conducted, it will be hard to ever "prove" that one house system is more correct (assuming that "proving" astrology is even something that should be done).
(2) Logic: One may believe a system is good or sounds right based on a definition or description of the system, or what looks right based on visual placement of the house cusps. For example, consider
Sinusoidal Houses for a system that has a simple reasoned definition, and also makes a case for being the "best looking" house system!
Logic can be subject to error too, because a system may seem right on paper but not actually work or be useful in practice.
E.G. WHY is Campanus more reliable than any other?
Every system has its own arguments, of course.
For example, Campanus being the "best" system is based on the following logic or beliefs:
(1) Houses are similar to signs, in that they're both 12 divisions of the sky. Since signs are always equal sized slices, houses should be too. House division should therefore be an Equal system (at least when viewed in 3D).
(2) Houses are inherently tied to the local horizon. That's why the first house is always near the Eastern horizon, in the area just below it. Combined with point #1, that suggests the best model for houses is 12 equal sized wedges placed over the local horizon. For example, it's important that a planet above the horizon will always be in houses 7-12, and a planet below the horizon will always be in houses 1-6.
The above describes the 3D model of Campanus, which is the only way to make the above conditions true. Campanus also has the "others use it too" reasoning, since it was the preferred system of Dane Rudhyar, the father of modern evolutionary astrology. As astrologer
Richard Brown wrote:
"My personal opinion is that Dane Rudhyar did more than any single other modern astrologer to drag astrology out of the cesspool of the Middle Ages fatalism and negativity to allow astrology to grow into the modern age. It was a mammoth and life-long task, which is not yet complete."