If the authors were using kg/m^-1 instead of kg/m^-2 with the given information (96.8 kg; BMI, 30.2 kg m−1; height not provided), the patient would need to be over 3m (nearly 10 ft) tall. This seems unlikely. If it were true, it would probably make for an even more interesting case report.

Peer review for less well known pay-to-publish ($1800) journals like this one is by no means a guarantee of quality: http://science.sciencemag.org/content/342/6154/60.full

kg/m^-1 -> kg/m (or kg * m^-1)

kg/m^-2 -> kg/m^2 (or kg * m^-2)

;)

> Although there's a popular simplified model that expresses an ideal weight as a function of kg/m^2, the weight/height ratio is indeed expressed as kg/m.

That's a non-sequitur argument: the OP wasn't talking about height/mass ratio. BMI is clearly defined and it is not weight/height ratio.[0] You might as well have said "I think the units for BMI are indeed kg/m, because Abraham Lincoln was born in 1809".

[0] https://en.wikipedia.org/wiki/Body_mass_index