I'm not against AI - at all. I do not owe a SW DC, but I like their products as much I like other DC from different manufacturers.
As a telecommunication engineer I can say you that there are 2 kind of self recovering failures that normally affects AI transmitters.
1) inadvertently change the channel playing with pressures (e.g for SUUNTO and probably Uwatec, and can be easily resolved coupling the trasmitter again)
2) being temporarily jammed by an electromagnetic noise source (non-shielded electrical motor - DPV - and/or some other sources - even not-shielded spark plugs). This would often result into a loss of transmission, until the signal/noise ratio go over a required threshold - for some brands it requires to put the receiver very close to the transmitter, because they lower the sensitivity in order to be more resilient to unwanted noise.
Transmitter tecnology is the same for what concerns the transmission frequency - because this is imposed from the environment.
The high pressure sensor can also fail, but this is usually a non-recoverable malfunction: once the pressure sensor fails, it will fail forever. so the problem is the same for an analogic SPG (I saw a couple of them start failing because salt cristals or oxidation on the axle rotating the hand).
I experienced personally (during the last 3 years) 2 dives with a temporary loss of signal from the transmitter and also 2 broken SPG, the first one did not zeroed, stopping at 50 bar, while the second required a gentle bang on the hand to have a reliable reading. The second one was mine and I opened and fixed it just removing salt crystals - the sealing was inappropriate and some water spilt in during the diving week.
Comparing the two kind of failures, I consider the SPG ones more dangerous, because they could signal you more gas than actually is inside the tanks - and this is how I discovered that something was wrong: I didn't expected thos values according my experience and calculation.
Just my 2 c.