First, admittedly this is a bit of conflict in interest, so I'll do my best to keep to the facts. That said, If anyone feels this post inappropriate I'll remove it.
Second, and mostly moot... my specialty has been primarily L-series. I use the Nissan CAS exclusively in those applications and have tested to 7500 rpm, with no trigger errors, period. However, as I mentioned previously, I'm opting to read only the 6 slot row.
Third, I've only recently been submersed into the RB installations so specific personal experience is limited. For this reason, I contacted Steve Taylor, owner of Wolf3D, to get a broader scope. In short, Wolf's plug-N-play's use both rows, counting how many slots are within each of the 6, to determine where cylinder one is. The implication is that the actual timing is a function of the 6 slot row. I couldn't say if the strategy (including the unspoken strategy) is any different than AEM. However, I have his assurance that Wolf has proven to be consistently reliable to 10,000 RPM, with a properly functioning, untouched CAS.
Fourth, 'reading between the lines', he defended AEM to some degree, with the statement that these CAS sensors and their wiring are getting old and the *occasional* issue should not be a surprise.