Try sending email or SMS. That's what most people are complaining about.
Simply dialing another number does not cause you to get a GSM channel. That doesn't happen until you establish the call or a data session. With a live call, you also have to produce noise on your end of the call because GSM will not use the channel to transmit audio that is essentially silence.
I'm guessing Australia's network works differently because I get interference simply by dialling/receiving.
Dialing a number will not create interference until it is transmitted. On most phones, that should not happen until you hit the "Send" button. Receiving text can cause interference too because your phone has to acknowledge the receipt of each packet in the message. Sending text should cause more interference than receiving text. Whatever the case, the channel access protocols are the same in Australia as anywhere else that uses GSM.
Sorry to hear that the fix did not work. I think Chris probably put capacitors on the output of the mic amps, right where the output is attached to your mic cable. That's a good place to start, but a better place to put the caps is probably on the input to the mic amp. The problem with that is that it can affect your mic's sensitivity and may also affect the frequency response. It depends on the topology of that amp. Also, care should be taken to minimize the inductance of the ground path back to signal ground of the amp. Feedthroughs on the board can represent 2 to 4 nH of inductance and a typical trace will have perhaps 1 nH per mm of length. It depends on the thickness of the trace. The smaller the trace, the higher the inductance. Anyway, the parasitic inductance of the board traces can affect the effective impedance of the capacitors and move their self-resonant frequency to a lower frequency. The caps were chosen so that they would have series resonance in the center of the GSM band, creating a virtual short to ground at that frequency.
Whatever you get for mics, good performance around GSM phones will rely on not only good shielding, but good rf bypassing around the active elements in the mic amp and good rf separation between shield ground and signal ground.
I don't know about how your mic cables are set up, but I'm assuming that they have a pair of signal wire (+ and -), a signal ground and a shield ground. If so, it's common practice to isolate shield ground from signal ground at one end of the cable or the other. If it's hum you're fighting, then you usually separate them at the mic end of the cable. If it's rf you're fighting, then it's common to separate them at the far end (not the mic end) of the cable. You might look into whether you can disconnect the shield from signal ground at one end of the cable or the other.
If on your cables, signal ground = shield ground, then maybe you can replace the cables and get better results, leaving shield ground separated from signal ground at the destination end of the cable.
Regardless, sorry to hear that the mods didn't work. I was hoping for good results with a simple fix.