I agree, it's not a defect. I think it's impressive from an original 1983 supply, with no parts replaced, all electrolytes are original. The ripple, as attached, on C1 is 1V, when at 3A to a resistive load.
The shape of the waveform caught my eye, and the phase difference relative to the C1 ripple. The 12V section ripple is much better, interesting enough. Perhaps the 2N3055/TIP31 on 5V is way tooo slow.