It’s relative, you don’t need to understand all and everything of theses parameters and colour science stuff to edit a picture. That’s a part of the philosophy of Adobe’s products, most lightroom users don’t care and don’t want to learn this. Of course this is helpful when it comes to photo editing, and you’ll probably make greater adjustment knowing it. And that’s the thing that make darktable “complicated” to some users.
In the UX field we often use the example of the car human-machine interface. You know how to drive a car, right ? But do you know anything about mechanics ? Do you know how a engine is working ? Probably not. And does it matters ? Probably not if you are not a Formula 1 driver ! Of course knowing some will help you taking care of your car and all, but that won’t make you a greatest driver. So yes I think it depends on the targeted users too, to what extend they’ll be ready to learn and adopt a new workflow and stuff. A product can’t fit everyone’s need. I’ll say that daktable is for “power users” of photography too…
I don’t thing UI have to be simple at the first sight, what’s matters the most is the mental model it propose, the learning curve of the software and affordance of the tool.
Simply put, the mental model is the idea we have of how something work. This mental model define what we expect of a product (not only softwares, but it’s really important in human-machine interactions) and how we’ll want to use it and the first sight. Every mental model is build at some point, and a mental model is based on beliefs more than facts. A wrong mental model is often the source the of bad-use of products. For example, saying a car will slow down and stop if I just ain’t acelerate is a wrong mental model that non-drivers can have (you also need to be on the neutral point). To me it feels like lightroom choose to mimic the mental model of a camera and others piture editors, naming parameters after the ones you find in camera, even thought it won’t fit exactly from the technical point of view. Darktable however, choose to go for a more “scientific and analog” way, even though it’s no the usual way of doing. And that’s not really a problem, it’s even one of his greatest strengh. The thing is that most people don’t want to change from one mental model to another or add a new one, that’s also why it’s difficult to change from a software to another. But these users usually consider that a photo is 50% the subject, 40% the shooting and 10% the edit, and won’t probably need / want to edit a photo heavily too. If you do corporate portraits, a 1h edited photo won’t pay more than a 10min one, clients will barely notice anyway.
The learning curve of Darktable is quite high in Darktable compared to Lightroom, but not that much compared to 3D softwares for example. The thing is that we accept 3D to be complicated more than photo editing, also because everyone can take a picture, not everyone can create just a 3D sphere… But Darktable has a great learning curve, modudule by module, then drawing mask, then parametric mask. You can get the most of it in 20-30h maybe ? You won’t understand 100% of the term but that will be enough to edit 90% of your photos with the most classical modules. That’s not much. I’ve spent 200+h on Adobe After Effects and I know maybe a third of the software (not even sure), and absolutly nothing of the video tech part thanks to the software. On the other hand using Natron for 20h you’ll just now how to export a basic image with 2 filters but you’ll know way better the composing work you’ve actually done. It’s a choice to take, and assume, and darktable do it right.
The affordance is kinda how a product is self-descriptive. The most famous example are doors handles : no handle means “push”, a handle means “pull”, no matter what’s written on top. On software there are classic statements on the UI : a round box means “1 choice”, squared box means “multiple choices”. A sliders is meant to choose between more than 2 values, curve alows specific ajustments on at least 2 parameters… And to that extend darktable is really great, and the UI gave all the controls to the user, no matter if you’ll have to use 3 curves and 2 sliders to be precise. You can be precise and the UI is showing this too.
Not everything have to work out of the box, what’s important is if it worth it afterwards or not. Darktable is more powerfull than lightroom for sure, more accurate and precise. There is absolutly no doubt in that. I think sometimes the only thing to do is to admit that yes it’s not intuitive for novices out of the box, yes you’ll need to learn some complicated things, and no you’re not forced too. But at some point, if you wants precision and control you’ll have too.
A GUI complexity is always subjective, and in the end only the ease of use matters really. Ideed users are not engineers, they are photographers in need of precise control, otherwhise they’ll be probably using something else. What’s complicated is to “step into it”, yes, like Blender, Natron, After Effects, Solidworks… Every precise and complete software at some points. Even Exels is complicated if you never used it… The one thing for sure is that you can’t make a software 100% easy (out of the box) and complete (powerfull).
A software is never too complicated. It’s too complicated for a certain use, or a certain user.
same with F1 car, a family sedan and a dump truck - all are “cars” and generally have similar purpose, but their UI/UX is waaay different (normal driver, even a mechanic would get lost in F1 car interface)