Modeling fatigue is challenging because two physiologically identical people may have very different psychological traits and thus different work capacities. Likewise, two psychologically similar individuals may have different work capacities due to physiological differences. Perceived exertion is complicated.
The exercise capacity model combines the many physiological and psychological factors that determine an individual’s work capacity. As of right now, the parameter values are hard coded. The engine does not change the exercise capacity based on the configuration of the organism being simulated. The physiological results of exercise will be different for different patient configurations because of differences in properties affecting gas dynamics and how the body responds to changes in blood gas concentrations, but the point at which the patient get “fatigued” and can no longer work will be the same across configurations if the exercise intensity factor is constant.
The power level of 1200W is used as a basis for normalization because that is about the most power you can expect a healthy young human body to sustain for a few seconds, but that number is somewhat arbitrary. Use it as a guide to determine the exercise intensity for the human that you want to simulate. For example, if you are simulating an older person who has a lower maximum work capacity of 1000W and you want to see the physiological result of exercising at 200W then you would enter an exercise intensity factor of 0.2.
I will use the calculator for the rower ergometer to provide another example. Assume that two people row as hard as they can for 500m. Someone who averages 6.1m/s will row 500m in about 1 minute 22 seconds and average about 650W of work over the event. A person who is able to sustain that work rate for 500m likely has a maximum work capacity close to 1200W or higher. A different person who takes 3 minutes to row the 500m is only averaging about 60W over the event. That person likely has a much lower maximum work capacity, maybe as low as 120W. (As an aside, I should point out that 3 minutes per 500m is a very slow rate on a rowing machine.) However, both of these people are exercising at an intensity factor of about 0.5.
Table 1 in the documentation is there to help guide you in choosing an exercise intensity factor. For example, running at 7.1m/s is pretty fast. For a healthy athletic person, that pace is an exercise intensity intensity of about 0.36, but for a healthy sedentary person that pace may be an exercise intensity of 0.6, and for an unhealthy older adult that pace may be an intensity of 1.0.
For any activity that you are trying to simulate, choose an intensity that is a fraction of what you expect the simulated person to sustain for a couple of seconds.
I hope that helps. I think the Kitware folks (or maybe someone else in the community) plan to add models that link the maximum work rate to the physiological state, but even then the psychological factors will be missing. Until the models capture all of the factors there will be some burden on the users to prescribe an appropriate exercise intensity for the given simulation case.