All good answers and comments so far. The question summary, "Conceptually, the lower the propellant mass fraction the better, right?" is correct if you abandon the assumption of a specified propellant and the propulsion system achieves the mission's required ∆V; but see the concluding paragraph below.
In addition to the mass ratio the rocket equation has specific impulse (essentially exhaust velocity, but for historical reasons velocity divided by Earth g) as a variable. For a specified ∆V, increase the specific impulse and you can decrease the mass ratio. As previously mentioned, this means more of the final mass can be payload, as long as the higher-Isp propulsion system's inert mass (engines, tanks, etc.) doesn't eat up all the gain in final mass.
Real mission concepts deal with this trade all the time: what is the required ∆V, and what is the best option for achieving that ∆V? I'm most familiar with interplanety robotic missions, and we frequently need to decide among cold-gas, monopropellant, or bipropellant systems. The bipropellant systems have the highest Isp, with NTO/hydrazine systems (various forms of hydrazine, such as MMH or UDMH) running in the 300-330 s range. Monopropellant systems, hydrazine again but no oxidizer, typically have Isp in the 220-230 range. And cold gas systems, such as pressurized He or N2, run far lower than monoprops. But each has its place, usually determined by the required ∆V. From the Isp point of view it sounds like biprop systems would always win. But biprop systems have somewhat heavier engines, two big tanks instead of one (which makes for a higher tank mass fraction), more propellant lines (pipes), etc., so their inert mass fraction is higher. They are more expensive, too! With current technologies, if the mission ∆V is less than ~0.7 km/s, the monoprop systems usually win out, with a propulsion system wet mass (inerts plus propellant and any pressurants used) that is less than for a biprop system. If the ∆V is really low, say a couple tens of m/s, sometimes a very simple cold gas system does best.
The conclusion is that there is no completely reliable generalization like "lower mass fraction is better." You must consider the objectives, such as the required ∆V, and characteristics like Isp and inert mass fraction of the available options. When you get the project managers involved, cost becomes a consideration as well.