The 3-section C5 is wired with two sections in parallel, making it a two-section part, 80uf and 40uf.
We adopted the habit of using a 500-Volt part with two 100 uf sections. The voltage on C5 can spike up to 470 Volts or more when the tubes are cold. Once they warm up and begin to draw B+ current it falls to the 410-Volt range. I'm convinced the 500-Volt part will last longer.
Back in the day, a 450-Volt multi-section filter would have a "surge" rating above the so-called "working" 450-Volt rating. A look at the B+ on C5 with a 'scope shows a sawtooth-shaped 120-Hz ripple waveform. A DC voltmeter will average out the peaks and valleys of this ripple, riding on the 400-plus Volts DC. If the peak value of that ripple waveform pushes past 450 Volts, that's an unhealthy stress on the capacitor. A DC meter reading is only part of the picture.
Rather than try to find the right white/blue wire on S3C, finding that wire at B43 on the audio board is safer. Simply pulling that wire from that hole in the pc board will disable modulation limiting. If that's what's cutting your mike audio, pulling that wire would make a difference.
Too easy to find the wrong lug on S3.
73