There is a reason for this. The digital panel is simply a PWM (DF) signal from 0-100% applied to the power control on the LPS. It has no feedback for how much light it actually puts out at a given %PWM
Every K40 system if different in terms of how much current the tube draws for a given Duty Factor control on the LPS-IN. The light output (power) is also a function of the current and the state of the tubes gas and that changes with age of the tube etc.
As the tube is used-up (a function of time and power level usage) the relationship between current and light output changes. Therefore the relationship between %DF (current) and the light it outputs decreases.
That means that you cannot count on a certain DF providing exactly the same power from system to system and from job to job on the same system.
This is why a meter is required to know actually how much current is being drawn. As the tube ages more current is needed to provide the same light output until finally the tube will not produce any output at any current. This is why it needs adjust-ability which you can get with a pot or a digital panel.
Geek pause: the tube is actually a negative resistance device whose gas under goes dissociation during ionization. So admittedly what I claim above is not totally accurate as the the tube acts electrically differently in pre-ionization vs after ionization.
I think the digital panel give users the feeling that it is precisely controlling power when it is really no better than the pot at its repeatability.
@Nedman 's case is interesting because the LPS and tube were fresh and more ideal than our machines are actually running at.