As an aside, I wonder why Innuos does not support AES as the others do.
I’m going to put on my business hat and remove my hifi hobbyist hat and say its a combination of they are probably playing to their perceived strengths. If they believe they know more about usb or that they believe it is an easier interface to monetize, then that makes sense. With one type of output to try to optimize, they are probably better at focusing development resources in one direction. Support and the resources necessary to do it well is also optimized...
Since the reclocker came up in earlier posts, monetizing the usb as a business trying to make money is far easier. If your product puts out a fantastic aes signal and your clocking is optimized...how many $4k ADD ON boxes can you build and sell to hifi enthusiasts?
Anyone ever seen an AES/EBU optimizer? Linear power supplies that take said optimizers to amazing new heights? Neither have I. There is an established standard with aes that no one evidently disputes. Additionally, Innuos is smart to focus on the interface most likely to be fragmented...so its easier to gain market share and penetrate it. USB presents that opportunity. DSD 1024? yep. PCM 24/3072? Yep!
Its easier to sell a product with a moving target, where you can OBJECTIVELY differentiate yourself from the competition. If someone says box A sounds better than box B, that can be debated. If some says "we are capable of 24/3072 resolution and our competition isn’t"...thats quantifiable, printable and repeatable but the secondary question rarely gets posed...does it matter?