From the article you linked: "Fleck and Falkel (1986) site a study that demonstrates compression forces are not excessive when the lumbar spine is flexed and the weight is held close to the body. In addition, intra-abdominal pressures generated during heavy lifts reduce extensor movements, thus further reducing stresses on the lumbar spine."
Compression forces aren't generally the problem, shearing forces are. Shearing forces may be reduced by holding the weight close to the body, but this does not show that lifting with a flexed lumbar spine is good for your back. When you flex your lumbar spine, your spinal erectors in the lumbar region (mainly iliocostalis and longissimus) are placed in a position where they cannot counteract anterior shearing force.
"Like you, I once believed lower back pain was mainly caused by a weak abdominal, flexibility, etc. until I meet a researcher whose team published their findings referenced below. The studies and accompanying papers strongly suggest other biomechanical factors contributed to the health of the lower back, specifically lower back strength throughout a complete range of motion."
In his book Low Back Disorders, Stuart McGill showed that neither great lower back strength nor increased range of motion were correlated with decreased incidence of low back pain. He actually thinks ROM is a silly way to assess back health. The factor that is correlated with decreased incidence of low back pain is muscular endurance.
"The avoidance of full range of motion promotes deconditioning and consequently deterioration of the joint structures. (Nelson 1993, 1995) This short sightedness is in total contradiction of the most basic principles of exercise: specific adaptation to imposed demands (SAID)."
Is the study referring to the zygopophysial joints of the spine, or joints in general? One of the main causes of a pars interarticularis fracture is repeated extension/flexion of the lumbar spine through its full range of motion. In fact, most spine injuries are due to repetitive actions of the spine and occur at the end range of motion.
Comparing this article to McGill's book, I'd have to side with McGill. His arguments make more sense and he uses relevant studies and anatomical evidence to support his claims.
I don't know much (anything) about mountain biking, so I can't answer your question, but it seems that your spine will be at least slightly flexed when seated on a bike because of the position of the handlebars out in front. In addition, there wouldn't be a lot of compression forces (unless your helmet is extra heavy) or anterior shearing forces while riding, so I wouldn't be paranoid about maintaining a tight neutral spine.