Welcome Einar...
great to see somebody doing "homework" before going Live!
I'd have to agree with your thinking, you're on the right trick... I might make these comments:
The H antennas /amplifier and the computer system, as it is presently processing, does not detect "Intra-Cloud" or "Cloud to Cloud" strikes... well, once in awhile, maybe.
This type of processing is pretty efficient at Cloud to Ground. It may be that many of the strikes you heard were CC or IC.
Actually, 4 stations will nail a stroke, if they all 'surround' it they'll do it pretty accurately. But I believe, generally, in the European region, the server wants 8. In North America, it's 6. And the system will use up to 12 for refinement. That may not happen at the split second they show on the map, however.
Stations too close to a stroke normally won't 'detect' it, simply because of the other 'cloud' noise. They may detect it, but it can be of little use for locating, etc, because of distortion. Signals can actually come out of the amplifier with clipped and distorted peaks. Now, if the station is running very low gains, it's possible. The closest stroke I've registered as located has been about 5km...
This is where the other network stations take over... The closest stations to the strokes, say within 30km typically will go into interference if there is enough activity, and won't sent data to the servers. This is simply to keep the servers from too much traffic, especially when most of the signals are likely to be unusable.
Now, all these issues are being addressed during the 'hiatus', as we've come to call it. The new E field antennas have 3 channels of available data, and that system can, potentially, detect the other types of lightning, IC and CC, and eventually we can get polarity, strength, and possibly altitude someday. The H fields cannot discriminate the IC and CC, and actually only work for CG. Furthermore, only one channel is selected by the server, though two are available. So the developers are working on many items with the infrastructure... including algorithms, auto gain control by the server to our controllers, etc. When they have all that done, or roughed out, they'll probably begin design work on the next generation of systems... I wouldn't look for 'major' changes in the 'kits', as far as function and capability, since the Reds with E and H fields are excellent, but I suspect the kits will be simpler, with 'physical' changes that will make them simpler to implement.
And your basic assumption of
"5 or 6" stations isn't enough" is absolutely true. The more stations in the network, the better the network performs. A perfect network would have stations gridded throughout the world 30-50km apart! One thing I always have to remember, my station by itself is just a bunch of flashing lights and buzzers and screwy waveforms on my signals page. It needs the other stations to actually be meaningful. And If I sit here with my gains turned up to get a signal from Jupiter, I'm worthless on Earth, because most of the data I send is just crap. So you'll see many stations gradually back down their gains, as the network grows, until eventually all of us are operating in a 30-600km radius.
Other strokes aren't detected simply because they are too weak to make it through the rain, cloud activity, etc, to enough stations. And you're correct about getting 'shouted at'... signals further than about 50-80km likely aren't "ground wave" signals, they are more likely "sky wave" reflections, and they can be strong enough and distorted enough to overcome and hide nearby signals... especially with weaker nearby energy.