You know, Don F's been exploring this somewhat. I've been chasing it 'mentally' since I went live last summer/fall, also.
Here are some caffeine deficient thoughts this morning....
1. If a network of stations were 'perfect'... all stations within, say 100mi of each other, then an excellent 'effectivity' would probably only be 20-40%.
2. Tobi tends to view 'effectivity %' as a general basic reference, not to be isolated as a measure. Now that's not strictly true, as I believe he'd tell you, but he tends to downplay its importance relating to station performance.
3. I'm currently thinking, which may change, that some type of interpretation of the "signals" total vs "EffectivityS" total for a given time, as a ratio, might be a better indication of overall performance, since it seems (I'm assuming, for now) to better express most factors, location relative to storm, station tuning, unusable signals. etc.... This is not a not-well-thought out idea - but:
...say Station X looked like this:
2014-03-29 12:58 9909 0 | 0% 1 | 100% 4538 | 72%
And Station Y looked like this:
2014-03-29 12:58 26664 0 | 0% 1 | 100% 5739 | 91%
Station X, then would be 4539/9909 or .46
Station Y would be 5740/26664 or .22
This might indicate that station X had a better tuned, better performing station overall than Station Y, even though Y had a much higher 'Effectivity' percentage than X.
If that were applied to a more dense network than we have currently, it might be one indication of "overall" performance. With our currently 'thin' network, where each of us is 'reaching' further for detection, and with my system currently dialed back because of nearby storms, and the Truck Axle Plant's line #3 Welder operating, it makes more sense than the "Effectivity" %... there is no way I'd ever achieve an overall "95%" effectivity compared to other stations. Likewise a Station in Los Angeles will never have a high Effectivity if all the storms are on Atlantic Coast.
Now, we all know there are other factors, but this might be something to consider. Also, especially with the new E-field receiver coming, there will be other data to consider beside 'strike number' and 'location ability'.. class of strike "C/G C/C" strength, polarity - possibly even "G/C"... who knows.
The sad thing, it appears that although Reds may be capable of accepting the new E data, I don't think greens will. So Then the network will have those other parameters to consider. As the network also improves in density, the need for 'longer range' detection individually will decrease. It's hard for me to remember that this is a "network" concept, not an individual station's performance taken alone.
But as you say, to contribute effectively to the network, we need some way to 'quickly' determine through the data available to us, if our station's parameters are at their best!
Mike