hi,
After making sound, according port pin must always return to high state
Kai, what does happen if due some reason (noise, power problem, program malfunction etc) a port pin goes to low level and keeps at this state for long time? For example, if MCU hangups while playing a sound? Maybe is it the reason to make scheme with some more transistors to make "middle dot" and capacitor-speaker-ground net from it?
BTW, depends on sound quality you need, it is possible to connect a capacitor between port pin and resistor to base of transistor. I have checked it and found that 1.5uF is good enough. But I would not recommend such scheme because you said right: the capacitor charge/discharge cycles are not symmetric due different internal n/pFETs of pin and so it produces worst sound.

Regards,
Oleg