Is it worth it I ask? Even if you could easily and efficiently bridge the two outputs you only (at best) get double the power - that's a 3dB increase in power and ditto for sound pressure level - equivalent to about a 30% increase in loudness. So, is it worth it?
You could use two 1:1 transformers - primary_one connects to +OUT_R and -OUT_R and primary_two connects in the same way to the left channel. Then you wire the two secondaries (anti-phase) in series and connect the ends to the speaker. You'll get twice the voltage out but you'll need to use an 8 ohm speaker to prevent overloading the chip. This means power doubles.
You could argue that the transformers are small and could be useful components in the EMC filter (needed for class D) but I don't think it's worth it.