I can give you an example that this is not always true though, from my own company.
We almost never do 16-bit color production, so we don't need that per se.
We do a lot of comps and "Freistellungen" (cutouts?), it's an operation that is required for 75 % of the output for print we produce. We mostly produce for web (no 16 bit colour anyway with most of the audience) and for really big advertising posters, up to 25 meter wide, some even 40 m wide.
The image production department ("art department"?) is managed by my wife. She used to hang on to Adobe products for the longest time, and she has had an art education. The problem was that she and the other employees of that department, were cutting out and compositing like it's taught in college, with the pen tool and a lot of patience and a lot of layers and a lot of time and a lot of degeneration of quality because of the enormous post-production. ProPhotoRGB and 16 bit were necessary, I used to shoot everything on a medium format digital camera to even have a good enough file to have an acceptable end result.
That's how pretty much everybody does it, and that's fine, but that's not a useful way to make money. The image production department was not making any money, and was not offering a competitive advantage, whereas all other departments were doing really well. The art department also was expensive, because Adobe software cannot access network attached storage like open source software, costs a lot of money on licenses, and has a lot of problems with funky updates and stuff like that, so the downtime is considerable. Add to that the huge downtime of the Windows PC's this software was running on, and it was a complete mess. Everyone else had been using open source software for years. I myself switched definitively to open source software long ago in university, back in 1996.
In 2012, I got the assignment of doing a large campaign, and for the first time, there would be the need to immediately respond with advertising material to themes that would pop up on social media... everything was super last minute, reaction time was crucial.
As a consequence, I shot small sensor cameras ("full frame", the small sensor size, not the much larger medium format sensors), because there were a lot of run-an-gun photos necessary to really connect with the actual super hot topics in social media, and medium format is really unwieldy for reportage-style shooting. This caused a drop in original quality of the files, enough to cause a drastic drop in post-production output quality after Photoshop.
An extra problem was the bottleneck in the art department. It was crucial to produce results in only a couple of hours, and the art department was taking 90% of the time budget, and everybody was unhappy and the entire assignment was very stressful.
To gain some original image quality, I processed the raw files in digikam manually, which was not really a time difference compared to the art departments Adobe Camera Bridge processing, but the image quality was noticeably much much higher. But the Photoshop bottleneck made me consider refusing assignments like that in the future.
Back in 2013, I reinvested in a completely new future-orientated platform for compute performance. Adobe software could not take advantage of that, it can't even take advantage of the multiple threads within the CPU of the system it is running on, let alone use shared resources. For everyone else though, this made a lot of sense, for video production, for development, for modelling, for simulations and for research projects.
I gave the art department a deadline to switch to open source software in 3 months, and sent them to full open source software training. Two people left because they didn't want to learn anything, the entire art department was not happy, but my wife runs the department and it was clear to her that they were not making any money there.
Even when the people came back from extensive training, they still didn't work efficiently, there was little difference between GIMP and Photoshop, in fact there was none, not in image quality and not in speed.
Until I made a challenge that the software developers and sysadmins in the company could outperform the art department in cutting out and compositing. Turns out that they could actually do the same work an art department employee did in 2 and a half hours, in less than 4 minutes with GIMP. The reason is that GIMP has much more evolved mathematical functionality than Photoshop. You simply don't have to cut out people with pen tools any more, you can do it very simply with preconfigured scripts and some simple logic. Photoshop doesn't have that functionality, you can simulate it, but it takes a really long time and several consecutive blurring operations that cost an enormous amount of time and deteriorate the image quality considerably.
Now the art department is also competitive, and the employees have learned how to make intelligent use of the GIMP, they use linux computers since 2014, so there is no system breakage anymore, they can work on the network with all the benefits, including access to the processing power of the ARM array, etc... everybody's much more happy.
And that's the mistake a lot of people make: they think the GIMP is a clone, a copy of Photoshop, and that it is not a very good copy or that a lot of functionality is missing. No, the GIMP is a very long running open source project, and a lot of people invested a lot of thinking into the functionality of the GIMP. In fact, probably more intellectual assets were invested in the GIMP than in Adobe Photoshop. As a consequnce, the GIMP has a lot of "level Asian" functionality in comparison to Photoshop. Art people don't like math and they don't like change and they don't like other people (especially nerds) telling them what to do, and they like brands..., yeah yeah, but companies like efficiency and profit, so art people have to learn how to become profitable, how to gain that competitive advantage.
In business, standing still is being left behind. Adobe and Microsoft have been practically standing still for 20 years. Of course it's convenient to not have to learn anything. that convenience also costs all profit and the little operational profit from the use of the Adobe and Microsoft products there is, is then shared with Adobe and Microsoft.... that is not progress, that is not good business, that is not economical, that is not intelligent.
What IS intelligent and good business and efficient, is using open source software, training people intelligently, so that they actually understand what is going on behind the GUI, so that they can - if they want - help in improving the products even further for the benefit of everyone.