Gotcha....
... yes, I'm not currently running an experimental vertical because of the noise.... but I've not really begun the experiment, yet.
In it's simple form, Blitzortung operates on TOA and TOGA of lightning impulses, so it doesn't particularly care what the polarity of a signal is in order to detect and locate a stroke. However, the system is capable of much more, and there is a lot that could be done with data from RED and BLUE systems. RED is capable of sending two H channels, and 3 channels of E data with 3 separate E channel bandwidths. BLUE is designed with 3 H channels, and only one E of wide bandwidth, using the optionally installed filtering instead of 3 E channels.
Since, most impulses are cloud to ground and are vertically polarized, the H loops and E probe network arrangement detects with good data quality the large percentage of strokes.
--- some horizontally polarized signals are of course detected incidentally, typically those nearby, but they are mixed with all the other sferics.
Now, by utilizing the first two BLUE signal channels, A&B, at 90° allows great 'omni' coverage of H 'vertical', same as RED-- and using the 3rd BLUE H channel, C, for a horizontally polarized loop, which is by nature omni, we could theoretically capture a few more strokes, and because of the data being on a discrete channel, separately track those impulses, if we wish. We could also discriminate better, by comparing, confused sferics that occur nearby, as well as distorted polarities and frequency distortions from reflected skywave components... (virtually any sferic signal > 50-100 miles would be a skywave reflection)
There is, then, no advantage to using 3 loops for horizontally polarized signals (effort is duplicated), some slight advantage using 3 in delta- for vertically polarized, and some greater advantage for using two 90° for vertically polarized, and one for horizontally polarized. (Once such algorithms are actively implemented on the processing servers.
Now, before somebody asks... please note that, unless very extreme shielded loops are used, the 'standard' vertically polarized H field loops, whether ferrite or wire, already detect E components of a Horizontally polarized signal! (Especially those nearby) So, because of the phase differences of the two, I suspect the server sometimes thinks these are H components of different strokes, hence many of the extra signals stations are (mis)credited with, in my opinion. Discrimination becomes worse with distance, because of the distorted skywave reflections, and is often worse with lots of active nearby cells...
So, in my environment, I'm able to discriminate the nominal vertically polarized E components with my E probes on both systems, but getting the extra near-field components eliminated from a vertically mounted H Horizontal detecting antenna will be difficult, if even possible, for me in my location.
Hope this gives some type answer... there are a lot of variables in this, and we learn as we operate... we already have situations where 'logic' and 'theory' play games with actual operation!