There is no shock, from any of the TK, Akula, and Ruslan devices, at the output. As far as I've seen, or felt on my own unit. As magnetic current is the main driving force, at the output.
   
   SR 193 had mentioned that the spark gap does not need to have an "intense spark", to do it's job. Although he only was able to light a 150w bulb, and not any higher wattage bulbs at that time.

    Both aquarium devices, use that type of bleed off spark gap. To drain excess charges, not to cause the spark. If there is no excess to drain off, due to higher loading, the discharge spark won't fire.  That's my theory, for what it's worth... Or why would you allow the spark to discharge like that and be lost? 

   Ruslan has shown a mayor spark when shorting out the rails on his 0.47uf tuning cap, but, no shocks from the output coil.  As it's mostly magnetic current, with only very little hot electric current, comparatively. 
  Akula's spark on his second video device (1500w) , was barely visible.

  The expensive 0.47uf caps used by Ruslan is the trick there. A similar WIMA 0.47uf snubber capacitor cost $30, each.  Non snubber 0.47uf 2000v (below) cost 12 to 15 dollars. And, Ruslan uses several of them in parallel. He is also now giving the Kacher a 53 volt input, he mentioned was needed to obtain the higher output effect shown in his last video, of 4000w output.
   http://www.ebay.com/itm/1pcs-WIMA-MKP10-0-47uF-0-47-F-2000V-5-pitch-37-5mm-Capacitor-MKP1U034707G00J-/161433959976?pt=LH_DefaultDomain_0&hash=item259636be28
 

   Building up a Kacher circuit that uses real high voltages may not be necessary, nor is there a magic frequency that works best. The spark gap is only needing to exert a high voltage at a matched frequency to trigger the proper heterodying effect.  Since the band width is very, very narrow, most people will not find it, or be able to tune for it, even guys with all the test instruments, like Itsu. He could not reach the right frequencies.
   
  Spark gaps are easier to tune, than electronic circuitry.  But, they are noisy, let off smelly ozone gas, and effect nearby electrical devices. Up to 1 kilometer away, according to Akula's third device tests.

  Ruslan said that if he takes his device apart, it won't work the same if reassembled by someone else.
   Well, I wish that he would let us "fools" be the judge of that. And just make a long instructional video, to show us all HOW he re-tunes it up, again, so that it works like before.

Its interesting how fierce the spark was when Ruslan attached the second array of lamps to the device in the 'forest' video.

Fierce sparking occurs when the current is smooth DC not AC, because the polarity change of AC (also pulsed DC) extinguishes the spark 100 times a second. With DC you can draw very long sparks once the spark has been ignited.

OK,  if you were deep in the woods and you would want to simply illuminate some 220V light bulbs, would you need an inverter? Not really. Connect 16 batteries each 12V in series and then you have 192V DC in total. The resulting brightness should be sufficient for a video camera. Hmmm ...

  The newer switching inverters can have a 90% efficiency, and weight a lot less than 16 12v batteries, and only need one battery.
   At 4000w output, like Ruslan is showing your 16 batteries won't last very long. Nor his diodes, as well.
  Unless you like carrying big car batteries around to the field.  I'd use a led camping light, instead.

  Reminds me of the Kapanadze unit that was taken from the sail boat to the beach, a long time ago.
    No piles of hay there, just rocks. And some very bright lights.

   Even a couple hundred watts output would do, to light their campground there on the beach.
It's not a question of how many watts output, but, that it was self running, and putting out a good amount of light, as well.  And there is nothing that indicates batteries, powering that much of a load.
You are again seeing, what no one else can see.

Using the right bulbs, is also very important. No self runner of the Akula/Ruslan devices has been shown self running, when using small wattage bulbs. Please correct  me if I'm wrong.

I'm a correcting...

When Ruslan first starts the device, it has the bulbs turned off and the system loops just fine.  Plus, there is another video where he simply runs an electric convection heater.  A resistive load is a resistive load.  Please correct me if I'm wrong.

Ruslan states repeatedly this device is simple to make run.  So we just have to make an honest effort to prove him correct.

Most interesting: At 9:09 Ruslan touches the antenna with a screwdriver so the high voltage draws some sparks. Although we can't see the light bulbs directly under the table, we can see their reflection in a cardboard box in the background. Astonishingly the brightness of the bulbs is in no way affected by the disturbance of the antenna. How can that be?

Because the device is allegedly very sensitive to tune, shouldn't touching the antenna detune the coil so the output voltage going to the bulbs changes a bit (the voltage reading on the multimeter does change indeed)? Just curious ...

Most interesting: At 9:09 Ruslan touches the antenna with a screwdriver so the high voltage draws some sparks. Although we can't see the light bulbs directly under the table, we can see their reflection in a cardboard box in the background. Astonishingly the brightness of the bulbs is in no way affected by the disturbance of the antenna. How can that be?

Because the device is allegedly very sensitive to tune, shouldn't touching the antenna detune the coil so the output voltage going to the bulbs changes a bit (the voltage reading on the multimeter does change indeed)? Just curious ...

Very interesting observation Zeit. I would certainly expect to see a change in light level. It certainly did whenever I touched my antenna.