It's not overblown, it depends on your use case. Octane has a near instant live viewer. You unpause the render and instantly see how your scene will look. You can move the camera and lights in real time. This is a huge time saver when working with lights and reflections or to just to check to see that everything is rendering correctly. With Iray, it takes a while for a render preview to occur. Idk how much Iray has improved over the past year, but for me it would take over a minute just to launch a preview in the Viewport, which was very laggy. Octane also handles low light conditions much better than Iray. A night time scene took me over an hour to render in Iray. That same scene took less than 10 mins in Octane. The difference can be huge.
Adding extra effort on every asset is a non-issue. It takes seconds to quickly adjust the parameters as Octane auto converts Iray material. The main effort is the initial investment of learning how it works. Additionally, Octane has its own database of materials that you can apply on any surface. Plus, this sort of initial tweaking is usually done once since most games set up environments once and then reuse them throughout the story.
Again, Iray is not bad. It's the ideal engine in Daz because all of the assets are created specifically for it. That's why I keep stressing that if you wish to continue using Daz as a hobby on the side, there's no point in switching to Octane or any other engine. If you wish to continue growing and advancing in the 3D world, you will hit some obstacles (like ridiculously long animation render times). Everything can be done in Daz that can be done everywhere else. The only thing other tools will solve is speed, efficiency and ease of use.
Quick example would be cloth sim. Dforce can accomplish some amazing things but it's slow and takes a lot of time to figure out the correct settings to get great results. Marvelous Designer cloth sims are real time and don't explode. It takes me less time to export into MD, simulate and import back into Daz than it would to do it all with Dforce. Depending on your goals however, it may not be worth the $500 cost of MD.
I'm simply explaining the pros/cons and effort required for the goals since Turitar asked about Octane specifically and mentioned that render times are a huge issue for him currently.
They updated iray for turing cards quite regulary, preview are quasi-instant, around ~15/20sec for heavy scene and you can travel free on photoreal mode without lagging. The gap with Octane is far from what it used to be, but I still think it's somewhere ~25% faster overall. And of course, if rendering time is your main problem, you should abolutely go for it or another engine.
That said, last time I checked autoconvert was quite a mess, you still have to fiddle shaders for nearly every surfaces (which can be a lot depending on scene/asset), and plug-in was crippled with bugs (like invisible meshes, and outright crash). Same could be say for Blender and diffeomorphic, solutions are far from being perfect and create problems of their owns (technical one or quality wise), a simple google search should tell you. And like you said, make assets re-use a little bit too tempting, which I find already problematic in lot of VN. More freedom for composition or iteration is always neat.
"Everything can be done in Daz that can be done everywhere else", true but my point still stand, I don't really see the benefit of switching to Blender or Octane for a kinda standard porn VN with a decent card, outside marginal rendering time gain or getting serious with animations. To "continue growing and advancing in the 3D world" (lol), I focus on zbrush & substance as I have no real prob with iray for what I do.
dForce always been horribly tedious and I think it will stay like that as they cornered themselves with "everything need to be physically correct" path. I also use MD for cloth sim when scene need it, that said a lot of casual clipping can been resolved with mesh grabber.
But overall I agree with you a lot even if it doesn't looks like it, in the end all that is matter is your scope and a decent workflow.
