Well, I agree. The extra cost of the mic is not a showstopper. In fact, for me the auto-cal feature would be the main selling point of the 3805. Without that, I would be content to pick up a reduced-price 3803 or even a 2803 and save a lot more than the $65.
So the effectiveness of the auto-cal process (especially the PEQ) is critical to my decision. So far, the reports in this forum have been quite favorable. But I have also read some of the criticism of the Yamaha YPAO system, and that raises some doubt about whether similar problems may be present in the Denon system.
One of the problems with YPAO was described in the recent S&V review of the RX-Z9 by David Ranada. His complaint is that all of the EQ modes available affect the higher frequencies, where they are primarily compensating for speaker characteristics, rather than room resonances and crossover problems, which dominate at lower frequencies. As a result, all of these modes resulted in unwanted treble boosts.
I get the impression that the Denon system doesn't work the same way. Still, there have been some complaints. For instance, the auto-eq sometimes gets the subwoofer distance wrong. Does this result in unwanted phase compensation? Small mains have been misidentified as large. Crossover frequencies have been automatically selected in places that people feel are inappropriate. And changing these settings manually after the auto-cal may not be the right solution: Does anyone understand enough about how the system works to figure out whether the PEQ depends on the crossovers? If the system is unable to distinguish small speakers from large, it might be feeding incorrect data into the EQ analysis.
(Theoretically, you could equalize each of the speakers over its entire range and then set the sub crossovers afterwards by hand. That is, if the frequency responses of the subs and the satellites overlap enough and can be flattened by the auto-EQ in the overlap region. That would make it a lot easier to set the crossovers. But in practice it might not work out so well, depending on the speakers and the room acoustics.)
I don't see how any such system could accurately distinguish between speaker behavior and room response, without requiring you to move the microphone around to compare near-field measurements with those from the listening position. Still, if the Denon system equaliizes low frequencies more aggressively than high, it may avoid the YPAO problem.
So I'd be interested in any clarification of these issues. Does auto-cal work the way it's supposed to, or is it something you have to work around to get satisfactory results?