Two in parallel per speaker for 6 µF in total? That is what makes sense.
3 µF @ 8 Ω = 6,631 Hz
6 µF @ 8 Ω = 3,315 Hz
My suspicion is that if this is a three-way the woofer and midrange are run full-range and overlap each other, and the midrange overlaps the tweeter. As I've elsewhere explained, the distortion from such out-of-band reproduction causes distortion, and adding a low-pass filter (inductor) for the woofer and a band-pass (capacitor and inductor) for the midrange will greatly clean up the sound. Also cover the midrange with a flowerpot, as is done for the Bozaks (extensively written about), to prevent cross-modulation between woofer and midrange. More the former influencing the latter, because the woofer energy is considerable.
I also suggest lining the cabinet with cotton batting, ala Bozak, to reduce reflections.
I would not rely on using leads as interconnection points, and instead suggest adding terminal strips. Very inexpensive and easily mounted on the board which appears to be ordinary hardboard (aka tempered hardboard or pressboard, common tradename Masonite), so nothing sacred about that.
The tweeters in those days were typically paper cones and poorly reproduce at higher frequencies, typically above about 6 to 8,000 Hz. I suggest investigating adding a super-tweeter, as has been written about for the Bozak. I have set forth a design which crosses in a soft-dome at 8,000 Hz which, in the case of Bozak, preserves the driver blending to that point and then runs to 24,000 Hz. The additional crossover is simple (each channel requires two inductors, a capacitor, two L-pads for balance, and the soft-dome) and inexpensive, with the cost being the super-tweeter.
Yeah, I know, I know, I can hear the critics lecturing me, "
oh, it's a console and the sound is expected to be poor, if something superior is desired scrap the console and buy real speakers", but I disagree: no reason exists to not spend a little bit on improving what could be improved and the improvement can be considerable. The original designers were aiming at a particular price point and the capacitors and inductors were expensive enough to make a unit far less profitable. That bean-counter decision — controlled by market forces — may now be reversed.
