I suppose it is conceivable that a suitcase-sized device could disrupt the actions of pacemakers of people walking within a meter or two. But it could just as well cause trouble with much more conventional means.
In addition, this emits a VERY wide bandwidth noise pulse (look at the frequency output of a spark gap transmitter) which of necessity puts very little power in any specific band.
That said, based on the size of that thing in the photo I suspect it wouldn't damage anything!
It takes a (much) higher fieldstrength to induce voltages that will destroy electronics than what's required to flip the occasional bit without damage.
Lightning can cause EMP like effects by inducing voltage in conductors not directly hit by the lightning.
its not linear.
either it induces a voltage high enough to cause disruption or it doesnt.
the limiting factor is the voltage not the power (power is tiny)
Vehicle disabling EMPs are of great interest to the likes of CHP, and such.
That has a chance of working due the very fast rise time you might be able to induce a sufficiently large voltage to cause some actual damage, it will be a 'one shot' affair though. But anything involving regular coils and some push button arrangement will likely cause it's own insulation to break down long before it will do damage to a phone.
That's one of the reasons the spark gap is there in the first place, without it you'd hit the limits of the coil insulation very quickly.
Spark gaps are a pretty nasty way to generate EM interference so there is some truth to the story.
But even with my plasma cutter (12.5 KW, so not the smallest in the catalog) and stick welder I never managed to crash a cell phone.
The plasmacutter did have the habit of resetting the controller until I properly decoupled all the lines into it.
Seriously though... can you imagine how many laws you'd break using that? The FBI would hang you up by your pubic hair, tar, and feather you.
http://www.rfcafe.com/references/electrical/near-far-field.h...
This thing operates in the near field, so it is pretty much useless past a distance of a coil diameter.
This gets around the square rule, sort of, and you won't need a suitcase full of batteries to power the thing.
That's what we used to test on stuff that was supposed to be reliable and it is amazing how hard it is to defend against it. If you could survive 10 seconds or more without a latch losing state or some other bitflip then it was usually a pretty good job. Even so, this required physical contact to something made of metal on the machine (even if most of it is properly grounded there always is some exposed metal that isn't).
I hope the engineering on that product is better than the English, it is barely understandable.
[1] http://gaming.nv.gov/modules/showdocument.aspx?documentid=29...
EDIT:
Holy shit, they have a video of what's left of their town after shelling stopped: https://www.youtube.com/watch?v=v4mc5mHksl8.
(Seriously, though, you'll need to square the power source as you double the distance.)
Or, possibly, explode into flaming death.