I'll toss out a few thoughts here...

The 990 does have superior bass management - four crossovers instead of two, and the bass management works with the multichannel analog input.

There are arguments that can be made for XLR being better than RCA - the arguments are most convincing for long runs, and many people suggest that on short runs it's not such a big deal.

The MMC-1's HDMI switching is no better than the 990's DVI switching (it doesn't accept audio and doesn't support transcoding from analog video inputs). In fact, the MMC-1's HDMI switching is not even as useful as the 990's DVI switching since the 990 will let you assign those two DVI inputs to any video input, while the MMC-1's HDMI inputs are permanently assigned to two video inputs (both of which are also permanently assigned to two of the three component video inputs). The 990's component inputs are also assignable. Picture this: you have two DVI or HDMI sources and two (or three) component video sources, and you want to connect all four (or five) through your processor. With the 990, you could assign any of those four (or five) sources to any of the six video inputs (DVD and Video1 through Video5) and then re-name those inputs.

The 40% discount on future processors is impressive, but I've got some concerns about Emotiva that make me question using the discount as a deciding factor in a purchase. I've been trying real hard to get Emotiva to offer some information on the myriad problems that exist with the LMC-1 - sent several e-mails and started a thread at AV123 where I took the time to compile a detailed list of the unanswered questions and existing issues (some very significant, I might add), and it took a total of six weeks for my five e-mails to elicit a brief "we're starting to look into these" response. The MMC-1 and DMC-1 are pretty solid processors, but they were derived directly from an existing platform (the Sunfire Theater Grand IV) so the engineering hurdles associated with surround processing, bass management, and the like were done. The LMC-1 is the only true example of Emotiva developing a new product on their own, and it has had repeated and significant problems. I also lost a lot of faith in Emotiva when they claimed that the LMC-1's DTS-ES implementation was fine and that all the other manufacturers were doing it wrong. Why buy something now specifically to get a discount on a future product (for which no pricing, specifications, or delviery time table exist) when an existing product has pretty fundamental problems like those described here ?