I suppose you could use 4P to make the 6-volt starter run faster. That is if the field windings and the rest of the starter can take the extra current. But they take 12 volts, so it would probably be OK.
You can do this, but it doesn't have the effect you're predicting... which is why Ford didn't do it.
If each field coil in the 2S2P arrangement has resistance R, the 4S arrangement has an equivalent field resistance of 4R, which is why it’s used for 12V conversions. But 4P makes the equivalent field resistance R/4, which is four times lower than stock. In a series starter motor, reducing field resistance reduces total circuit resistance and tends to raise current draw. How much the current rises depends on how large the armature resistance and battery/cable voltage drop are relative to the field resistance, but in practice it often rises a lot and stresses everything in the current path.
At the same time, the magnetic field per pole usually gets weaker in 4P. Field strength per pole is proportional to ampere-turns (turns times the current through that particular coil). In 2S2P, the line current splits into two branches, so each coil carries about half the line current. In 4P, the current splits four ways, so each coil carries about a quarter of the line current. So for the same line current, 4P cuts the field ampere-turns per pole in half. The line current will rise because the resistance is lower, but it usually doesn’t rise by a full factor of two because the armature resistance and battery sag are significant, so
net field strength still decreases.
That matters because torque is roughly proportional to field flux times armature current. If you weaken the field flux, the motor tends to run faster, but produces less torque per amp. Under load it tries to make up the torque by drawing even more current, which increases heating and brush/commutator stress. A starter wired this way will
appear to run faster when unloaded or lightly loaded, but it won't crank better under compression.
There’s also a saturation problem. Starters are designed so the iron in the field coils is near saturation at normal cranking currents. That means increasing field ampere-turns doesn’t increase flux very much beyond a point, but decreasing field ampere-turns can reduce flux noticeably. So 4P is more likely to reduce useful field flux than it is to unlock extra torque, and the extra current mainly becomes extra copper loss and voltage drop rather than extra cranking power.
