Theoretically, this would work – if you sit an ATX PSU in the bottom-rear, it fits within the enclosure and the height of the bottom four PCI slots. I suspect there may be some interference with the bottom edge of the board, however.
The reason we haven't supported this is twofold – one, it would add more parts, complexity, and cuts to the rear panel and associated brackets/plates, which would make the whole rear solution more cumbersome for everyone. Two, in our opinion very few people should be considering ATX power supplies given all the fantastic SFX and SFX-L options in the market today, and in their use case the top mounted location tends to make more sense anyways (mainly to open up the PCI slots).
Per the OP, our plan is to provide solid panels as the base, with the option for either a windowed side panel or a vented side panel. But solid panels all around is what is 'standard' (and in my opinion is what looks nicest!).
I will let [USER=10]@Wahaha360[/USER] answer with a more specific radiator width measurement since I don't have measurements handy, but if you look at the third render in the OP of CX2's frame, you'll notice that the rails for front-mounted fans are offset towards one side - this can actually be flipped in place, allowing the user to decide which direction the rails are offset. We did this deliberately such that folks can shift radiators and fans away or towards the motherboard as needed to ensure a good fit. In our experience most motherboards and radiator/fans work; I have tried various motherboards and radiators and have yet to find an incompatibility in that dimension specifically.
Given the front-to-back cooling design, and our intention to hit a competitive price (holes are expensive!), we don't plan on offering a top vented option. Our expectation is that it would provide negligible benefit thermally, worsen dust, and could actually diminish thermal performance in certain circumstances. (For the original Cerberus, the top fan didn't do much in our testing, but we provided it since it was 'free' to include with the optional handle bracket).
Just to double-down on it: we are firm believers in the front-to-back cooling design. Dual fans at the front are very effective at evacuating hot air quickly, and when investing in good quality fans, they can do this while maintaining reasonable noise levels. Although we are working on a vented side panel as an option, we think this is best suited for very specific builds where supplemental side intake/exhaust makes sense.
Per the OP, we are planning on opening up pre-orders on March. We want the design to be 100% finalized, with a prototype and price points to show for it, before we start taking folks' money.
We've been trying a lot of different approaches with the front panel. This has been perhaps the most time consuming part of CX2, actually; given that the enclosure is so boxy, we are limited in what we can do aesthetically to make it unique and beautiful. This is then compounded by the fact that the front-to-back cooling design requires that we create a front panel that adds as little airflow restriction as possible.
The unit that OptimumTech has is a prototype that has a front panel design we first tried out in October. Here are some glory shots I took of it at the time (I think it looks a bit better in silver than black, it's a bit reminiscent of the cheese grater Mac Pro):
[ATTACH=full]387[/ATTACH]
[ATTACH=full]388[/ATTACH]
You will also notice that the side panels extend beyond the front face a touch, and the corners are notched at the top and bottom. I actually quite liked this iteration. The current renders, comparatively, trade this for a metal mesh, and have the top and sides extend out front, with only the bottom corners being notched. This is our current preferred design since we believe it looks better and more distinctive, and would likely provide even better airflow. Naturally as we prototype and experiment we will share that progress with you all!
One other thing - taking a step back, one of our goals with CX2 was to support modularity as a feature, and to that end the front panel we ship on day one will not (in the lifetime of the enclosure) be the only option for end users. We may (for instance) offer this current iteration as an accessory, or experiment with other ventilation patterns or materials.
A quick note on the handle: Per the OP, our plan is for the handle to be an option, with the default being a solid panel with no handle. In the absence of the handle, there is ample clearance at the top of the enclosure, and all of the compatibility issues he mentioned go away.
OptimumTech didn't make the optional aspect of the handle clear in the video, but honestly that is on us because our coordination with our manufacturer got a bit confused, and ultimately Ali was not sent the solid top nor a side panel. As was presented to him, the handle was the default, and he had no way to build inside CX2 without the handle option installed. The video still turned out great; we were very happy with it and that's mostly thanks to Ali, who asked many good questions and ultimately made the video an opportunity to provide feedback rather than perform an extensive review (which would be rather pointless given that we are still iterating on the design). Needless to say, we were grateful for his thoughts and time.
Most importantly, in the video he talked about some compatibility issues with the handle that he encountered, and that we had not in our own testing. We are already reviewing how we can modify the handle to address these issues. IMO this goes to show how important diverse component testing is, and how fresh eyes and builds can always uncover new compatibility problems! Needless to say, designing for SFF is not easy; there is good reason that you see so many projects on SFF.N and such emphasis on community feedback for these products.
The handle has been an interesting feature for us to design for. It was important that we provide an option to support it, but at the same time we didn't want to compromise the enclosure for the majority of users who don't end up purchasing one. With the original Cerberus/X, I remember the many iterations we went through on this one feature, and we ultimately settled on a machined handle attached via bolts that slid through holes in the handle and top panel, into a bracket that secures to the internal frame. This worked, and (in my opinion) was fitting of the design goals of the Cerberus/X, but it had some significant compromises: the handle was expensive, and it would stick out permanently, which was a bit divisive from an aesthetic point of view. The handle for CX2 as it exists today is more economical and more integrated, but restricts compatibility for those who use it in ways the original did not. Particularly with what Ali found in his own testing, it's clear that the compatibility issues are too pronounced currently, and so we are investigating what can be done to mitigate them.
I do think this comment pertaining to the handle is worth highlighting:
We actually went through essentially all the same thoughts internally! For instance, our first design rev had essentially a "hat" that added an inverted, squared-off U to the top of CX2. This made CX2 taller and created an air gap above the top panel and below the top of the "hat", where one could grab the enclosure from the front and back. Ultimately we had to ditch this design since there was no easy way to integrate it into the frame in a way that was simple, optional, and cost effective.
We also considered things like straps and rails, but encountered similar issues, or decided that we didn't have confidence in the robustness of the solution in the long run or for heavier builds.
