The number everyone quotes traces to one paper: Firoz and Graber, "Bioavailability of US commercial magnesium preparations", Magnesium Research, 2001, 14(4):257-62. It measured how much magnesium volunteers absorbed from four products and found magnesium oxide came in at 4% fractional absorption, while magnesium chloride, lactate and aspartate were significantly higher and equivalent to each other. That is the basis for treating oxide as filler, and it holds up.
Two things that paper does not say, and that you will see claimed anyway. It did not test glycinate at all, so it cannot be the source for "glycinate is the most absorbable form", and we have not found a study that is. And its own conclusion cuts against the tidy organic-versus-inorganic story: the authors wrote that inorganic magnesium salts, depending on the preparation, may have bioavailability equivalent to organic ones. Oxide is poorly absorbed. That is a fact about oxide, not a ranking of everything else.
The reason brands blend oxide in anyway is arithmetic, not chemistry. Magnesium glycinate is about 14% elemental magnesium by weight, and magnesium oxide is about 60%. To get 200mg of elemental magnesium from glycinate you need roughly 1.4g of compound, which is a big capsule or a scoop of powder. From oxide you need about 330mg. Oxide is how you make the dose fit and the price drop, and the panel is where it shows up.
On the other forms: citrate is absorbed well and is the one with a real laxative effect at higher doses, which is useful if that is what you want and a problem if it is not. Glycinate is the gentle one on the gut. Beyond that, be sceptical of form-specific promises, including ours: the trial evidence comparing these forms head to head in ordinary people is thin.