Devs aren't complaining about it because fragmentation at day at the start of a gen hasn't happened -- as it pertains to performance -- and even with the differences in HDD size and availability at the start of last gen, developers DID complain, which is why we saw a much simplified offering with the launches of PS4 and XB1 at the start of this gen.
You seem to me to arguing from the perspective that, "MS is doing it therefore let me conjure up every contrived justification as to why its a good idea".
I'm arguing from a perspective of whether its a good or bad idea based on the pros and cons to every stakeholder, i.e. platform holders, pubs/devs, consumers.
So far, i only see pros for MS. Consumers get a semi-raw deal, as while they get a console that's cheaper, at the XB1X level it simply isn't next-gen, while for devs it quite an adjustment and additional hassle for development--which whether you want to believe it or not, will appear in the games.
Conceptually, releasing two skus at different performance levels is a worse idea than launching a single SKU and keeping all console owners on a level playing field... which incidentally was the entire point of consoles from the very moment they were conceived.
Anything that disrupts or adds complexity to that fixed hardware spec. is negative, from a dev perspective.
Anyway, you seem intent in dismissing these points, so there's little point in continuing this line of discourse.
I have the same concerns regarding a muilti sku approach as you've listed above. But, I think there are more pros to the situation than what you've looked at. There's also enough unknown information to cause our speculations to potentially be wildly off course.
Taking a step back, the first thing to consider, is that any and all pros/cons to this situation will have been researched by MS, backed up by large amounts of data none of us have access to. They're also likely to be in direct contact with devs, taking their feedback on these things. That's not to say companies can't make mistakes, but I think with the leadership at MS and Xbox, they are very aware and focused on getting things right.
Having to release a game on an additional sku does add complexity to the process. But, the degree of added complexity is something we don't know and therefore can't quantify. There are also pros to releasing on additional skus, that might outweigh the added complexity.
In the Brad Sams thread, Elenarie - a developer at Dice, commented that it
wouldn't gimp next gen to have a lower powered sku and was
positive regarding multiple skus.
A two sku strategy will likely be easier for devs next gen, than currently. First, I expect the two skus to be much more similiar in terms of architecture - having all components be similar, but just more in some cases on Anaconda vs Lockhart. Instead of today, where the memory system is different between the two platforms, minor architectural changes to the gpu, etc.
Secondly, the other part of the recent rumors is MS's work on Game Core, which seems to improve developer tools and make transitioning games to different platforms much easier.
Launching with multiple skus will allow MS to hit several price points targeted at key consumer groups. It's likely they're *targeting* to have both the least expensive and most powerful consoles next gen. This is a benefit to consumers because they have a choice in how they want to enter the new generation. For developers, they'll continue to develop on two Xbox skus, but the next gen platforms will be much more similar in terms of design - potentially meaning developing on 2 skus next gen is easier than it currently is. With their games launching on a platform with a more accessible price of entry, it means higher potential sales of their game, which is always a risk when launch on a new platform.
Just to be clear, I'm not sure if two skus is a good or bad idea, I just think with the right plan in place, it could be.