It makes no sense at all.
Thrust of a rocket can be defined as follows:
F = dp/dt
p = mv
m = mass of propellant
v = ejection velocity of propellant
t = time.
Usually, it is the velocity that is of interest to us. However, it can be said that within the operating propellant flow rates of the rocket, the ejection velocity of the propellant is more or less constant and defined by the operating principles of the engine.
For example, ejection velocity from chemical propellant rocket engines is defined by (oversimplifying a bit) the chemical potential energy in the propellants. For reference, see
Wikipedia article about liquid rocket propellants. The flow rate of the propellants affects this comparatively little, so instead of the typical expression
F = m dv/dt (or F = ma)
we can use the following form:
F = v
e dm/dt
where v
e is an approximately constant ejection velocity. Of course, there can be some variation at different thrust levels depending on engine design - typically they have an optimal performance range which they are designed to operate on.
Further inspection of the term dm/dt reveals it is basically the flow rate of the propellant - how much mass is ejected out of the engine per unit of time.
This, essentially, means that
within their operational range, most rocket engines have a linear relation between thrust and fuel consumption.
Of course, most rocket engines are also unable to operate at arbitrarily small thrust settings, when the minuscule propellant flow basically just scatters out into the combustion chamber, not generating sufficiently high concentration for ignition. While the small amount of propellant flowing through the nozzle would in fact provide small thrust, this would be extremely wasteful from energy perspective, as the ejection velocity would be very slow, and none of the chemical potential of the propellants would be used at all. It would be about as effective as a constant slow leak on the tanks.
Hypergolic rockets of course require much less fuel flow to function due to the volatility of the propellants, making them ideal for RCS thrusters and in-flight course corrections; the Apollo CSM engine for example was a bipropellant hypergolic engine built for reliability.
Ion engines, of course, would be scalable by retaining high voltage, but regulating the propellant flow to arbitrarily small values. This would be reasonably simple to do, and would provide a fully linear thrust response curve.
Ideally, chemical rockets in KSP should have a set "lowest possible" thrust rate enabled at 1% throttle, an optimal burn rate at around 75-80%, and maximum power at 100-120% for emergencies.
But they haven't even introduced fuel/oxidizer division yet; every engine acts as a monopropellant - which considering kerbals I wouldn't really be surprised by them using some insanely toxic and corrosive stuff as their main rocket propellant...