Pin 3 could totally be audio, we noticed that when the Tamago was beeping, the other signals were distorted, so maybe it was the result of crosstalk. I have several extra Tamagotchi noise-makers that have become detached in various accidents, so I might try driving one with this signal.

Haven’t been able to trace the power and ground to the figure contacts, unfortunately.

Would totally be intersted in trying on a logic analyser

Can you zoom in on the Pin 7 trace? Increase your sweep frequency and trigger as soon as it goes high. I would try the same for Pin 3. When I see whacky burst traces like that it’s often because I’m sampling too slow (below Nyquist i.e. < twice interesting-signal-freq) which leads to aliasing and a misleading or malformed trace. The odd-waveform on Pin 3 at 3+kHz is in the audible spectrum, so maybe this tamago is “talking” to you. You could hook that signal up capacitively-coupled through a pre-amp to a speaker to hear what it sounds like. Also consider the whackiness may be caused by your observation, i.e. the load of the scope probe. Ensure it is set to 10x impedance.

Pin 6 definitely looks like data.

Did you identify power and ground from your previous teardown?

Since all these signals appear to be 0-3V we could put all the lines on my Saleae Logic analyzer and capture them all simultaneously. Let me know if you interested.

Great work!
DW