skotdee, if your R-44 is like mine, then based on what I measured, I think you'll have no problem at all with it powering four of your AKG microphones. Doug Oade's summary is a bit too warm and fuzzy--quite a few microphones draw more than 4 mA, not just a few rare types--but if your microphones draw less than 3 mA apiece then there really is no problem.
kbergend, perhaps the issue is open-circuit voltage versus voltage under load. For 48-Volt phantom powering the specified tolerance is ±4 Volts, but that's open-circuit voltage--what you would measure at an XLR socket that has no microphone connected to it. The usual circuit feeds the DC to each microphone through a matched pair of resistors specified at 6.8 kOhms each for 48-Volt powering. When a microphone is connected, half of its supply current flows through each of those resistors--and the more current that flows through any resistor, the more the voltage drops across that resistor. So the voltage actually delivered to the microphone is less, in proportion to the current drawn.
If a microphone draws 2 mA, then 1 mA is flowing through each of the 6.8 kOhm resistors, causing a voltage drop of 6.8 Volts across each resistor. Thus if the open-circuit voltage was 48 Volts to begin with (and if the supply itself is perfect and doesn't sag at all under load), then the voltage actually reaching the microphone will be 48 - 6.8 Volts, or 41.2 Volts. That is completely in accord with the standard; only the open-circuit voltage is specified, and the actual voltage reaching the microphone is supposed to be lower in proportion to the current being drawn. The microphone designers know this and work with it.
If a microphone draws ca. 7 mA then it is really receiving only about 24 Volts. To draw more than 7 mA is actually to pass the point of diminishing returns as far as power is concerned--at 10 mA (Earthworks), the drop across each resistor is 5 x 6.8 = 34 Volts so the great majority of the power is being expended in the feed resistors, not the microphone circuitry; it is simply wasted as heat.
Now as it happens, the supply in the R-44 isn't quite perfect, and it does sag as more current is drawn--if you connect 1, 2, or 3 microphones at 4.5 mA apiece and measure the open-circuit voltage at the fourth, unused socket, it keeps getting lower as more and more current is drawn from the other sockets. Ideally that should not be the case, but as long as the equivalent open-circuit voltage stays in the 44 - 52 Volt range and the recorder isn't harmed by being pushed up to its limit, then everything should work fine.
Unfortunately what I found is that my R-44 can't actually deliver even 16-18 mA without the voltage sagging below the tolerance limit. Thus I avoid connecting four Schoeps CMC-series 48-Volt mikes at the same time, simply because the recorder is at or slightly past its actual limit then, and I'm averse to operating equipment that way. I have seen a phantom power supply literally burn itself out; it smells funny, it can be expensive to repair, and it spoils the recording, which kind of defeats the purpose of my being there.