According to the specs, the GM 3 bar reads the following (pressure is absolute):
.501 - .761 volts (.631 average) @ 40 kpa
4.784 - 5.044 volts (4.914 average) @ 304 kpa

The range is likely tolerance for temperature changes affecting the sensor, so I've elected to split the difference for the 40kpa and 304kpa specs and use the average voltage they can output. That leaves us with the following sample to calculate the scalar and offset with.

Pressure range: 2640 mbar (304 - 40 then x 10 to get mbar from kpa)
Voltage range: 4.283 volts (the 4.914 average minus the .631 average)

These two points tell us the scalar, which is 2640 / 4.283 =~ 616. Think of the scalar as "how much pressure each volt represents." The only thing left is the offset, which tells us "where the scalar starts at." Think of it like a linear function of P = (V * scalar) + offset where P is pressure and V is voltage.

We can figure out the offset by testing a known point, such as 3040 mbar (304 kpa) at the 4.914 average volts.

3040 = (4.914 * 616) + offset
3040 = 3027.024 + offset
12.976 = offset.

That makes the settings:
Scalar: 616
Offset: 13

As you can see, this is very close to the neetronics 3 bar map sensor profile, which is why it works well, but you should note the following differences in map sensor readings between the two sensors as boost increases. This is probably why people's boost gauges are off from the map at higher boost levels as well.

GM
616 scalar
13 offset

Neetronics
625 scalar
42 offset

@ 2.5 volts
1553 mbar 22.52 psi (7.82 psi boost) gm
1604 mbar 23.26 psi (8.56 psi boost) neetronics

@ 4.5 volts
2785 mbar 40.39 psi (25.69 psi boost) gm
2855 mbar 41.41 psi (26.71 psi boost) neetronics

So as you can see, at around 8 psi, you're almost a full pound (.75 psi) off, and at 26 psi you're a full pound off.

Obviously if you're going to try these settings, note your tune will be off afterwards.

Wayne