This is an excellent point. One potential point that could be in favor of series hybrid however is that if the engineers know in advance the engine only needs to run at a fixed-speed, it would be possible to design it with a greater max. thermal efficiency than a normal 4-stroke engine, which is great for torque and throttle response. Use of an exotic engine cycle, a recuperated turbine engine, or even some sort of external combustion engine might be a more efficient primary powerplant.

However, one advantage that the parallel hybrid system has over the series-hybrid is the ability to combine engine power and electric power if needed. Considering that many electric motors have excellent power density however should mean it's possible to throw a large motor on the rear axle, and one each on the front wheels to produce more than enough horsepower in a series setup.

Another idea I've thought of however for a series-hybrid is that a fixed-output engine could still be used, but with a single mechanical gear on a clutch that would allow it to act as a parallel hybrid at high speeds. If say 20 horsepower were needed at the wheels to cruise, and the engine were putting out 30 horsepower, Instead of converting all of that to electricity and back, it could be clutched to the wheels, and a 10 horsepower charging load added to it, sending 20hp directly to the wheels, and the rest to the batteries.

Finally, an advantage of series over simple elecrtic is that for short trips, the series model can act as a pure electric. However, the engine can come on for longer trips (sparing the batteries from being deep cycled, which should extend their lifetime) and if cabin heating is needed (electric heating costs money and reduces range, unlike using wasted engine heat).