While I'm no expert on either, having seen a few product demonstrations of both could you explain why you feel Clarisse is equivalent to Katana? I understood Katana to be a renderer agnostic lookdev tool. Clarisse has it's own high performance renderer and seems to have a broader scope.
Clarisse is like a cut-down version of Katana with a built-in renderer.
It's also got limited layer-based compositing functionality.
It's cool in itself, but it's a bit like a "poor man's" Katana in terms of what you can do with it from a high-end pipeline perspective, and Katana can scale to huge scenes which I'm not convinced Clarisse can (this isn't including replicators or instances, which are cheap and easy to do and both cope with this) - I'm talking unique geometry in the billions of polygons.
Generally the big studios will connect Katana to PRMan or Arnold to do the heavy lifting at render time.
They are both procedural applications for describing and editing scene descriptions, which allows for very powerful workflows, in terms of overridable attributes which propagate down the scene graph, and they both allow lazy evaluation, meaning you don't open the source asset files until you need to.
But Katana's got awesome recursive procedural support, which linked to PRMan and Arnold, allows literally rendering of infinite scenes, as well as the CEL language which lets you script nodes in terms of behaviour and modification down the scene graph, basically allowing you to do almost anything in terms of defining nodes, or attributes, items in the scene graph.
Basically, Katana's a huge toolbox that the big VFX studios are using to fit (some of) their pipelines around and write their own tools for.
Clarisse is more like an off-the-shelf out-the-box tool which can do some cool stuff which is useful to small teams and maybe independent artists, but pales in comparison to Katana in terms of capabilities.
Ok, interesting to hear your perspective. You are a Foundry developer though, right? I'd have to give the Isotropix guys the opportunity to make their case as well.
Of course, and they'd rightly say that Clarisse is useful out-of-the-box, whilst Katana most definitely isn't - you need a renderer, it needs a lot of pipeline work to integrate it with your asset management system, and it needs a lot of customisation to make it "nice" for artists to use.
Out of the box, I wouldn't even class Katana as a good look-dev tool, as the default light manipulators are pretty poor - most studios generally customise them and make their own versions for artists. But that comes back to my original point (which I deviated from quite a bit), which is Katana is more a toolbox with very impressive (but not perfect by any means, there are issues with it) infrastructure capabilities, that VFX studios can build on to fit into their lighting and rendering pipeline.
