While the magma won't shoot straight outward, a wide enough array of pumps will make it form a large glob that extends several tiles beyond the end of the barrel. Essentially, any pressure you have will cause the magma to expand in
all directions.
In order to work reliably, though, a simple gravity feed probably isn't enough since magma likely won't fall fast enough to keep all of the pumps constantly primed (haven't tested this with a uniform magma tank with the lower pumps submerged, just the bottom of a magma tube where the pump intake tiles were obviously underneath a ceiling and thus could only fill from the sides), but if you have a tall enough stack of pumps, it'll provide a decent radius burst.
Pressure applied by screw pumps isn't the same as water pressure - it only applies 1 "unit" of pressure whenever the pump moves an amount of liquid (meaning if you try to push 4/7 of water into a narrow that already has 5 tiles of 7/7 water, it'll get pushed all the way to the 6th tile), and when there's nowhere for the liquid to go, the pressure simply does not accumulate.
As a test, I built a 5-wide 19-tall cannon using the following layout, and the edge of the magma reaches 5 tiles beyond the end of the barrel before the upper pumps lose pressure and the flow reduces to a trickle. A sufficiently wide and tall pump stack, when primed, could probably flood an entire region tile in a matter of seconds, but it would require an ungodly amount of machine power.
Z+0
######
#_%%.#
#_%%.########
#_%%........X
#_%%.########
#_%%.#
######
Z-1
######
#.%%_#
#.%%_#
#.%%_#
#.%%_#
#.%%_#
######
Z-2
######
#_%%.#
#_%%.#
#_%%.#
#_%%.#
#_%%.#
######
...
A movie of this cannon in action can be found
here.