I couldn't really speak at length about Unity at the time but I'll contribute some
Unity has excellent potential for single man teams but has two main factors which are highly problematic:
1. High barrier of entry
Many titles seem to have a maximum on how much they can start with, and considering that it's been used so much by so many developers it stands to reason that with Github and such there would be even more of the aforementioned lame games. Usually that produces a high quality game as they start to see the titles they're competing with.
2. Low ceiling
Similarly it lacks the growth capability when you have a larger team, and need to save time optimising and manage gameplay (don't have time to elaborate but this is the best I can do)
I'm really not sure I agree with either point here, if I'm honest.
Unity has *a* barrier to entry. Of course it does. It's a tool that requires some investment of effort. I have yet to see a tool that would give you meaningful results that has zero barrier to entry.
It doesn't even matter which arena we look in, there are barriers to entry, even in the tools that provide a lot of handholding, and the more handholding, the more ceiling there is.
You're not making anything other than a JRPG-style game in RPG Maker without a significant effort investment; you're making a conventional JRPG game with *some* investment. But it still requires an investment of learning, though much lower than Unity.
Getting more fancy, AGS, Visionaire, AGI or SCI Studio (god help you), and friends, all let you make a certain style of game (with some getting into other things if you really want) but again with a barrier to entry.
Over in other corners you can start talking about things like Pico8 and its particular relatively low barrier to entry but again low ceiling.
Everything's got a barrier to entry - and I don't believe Unity's is significantly higher than any other tool in its weight class; it's not radically more obtuse to get started in Unity than it is in Unreal, or Godot for 3D stuff. For 2D I might argue that GameMaker Studio is slightly easier to get going with, but not hugely.
I will give you that there is a huge discrepancy between the ambition and the execution of many wannabes who buy into the idea that anyone can make a game and that Unity gives you all the tools. It does - but you have to learn how to use them.
The comparison with Photoshop is apt; it is an utter powerhouse. Vastly more powerful than most other tools out there, but you have to learn how to drive it. Nothing is going to change the fact that you have to invest some time with your tools to be able to use them well. Now, there are tools that for some uses are easier to drive and better suited for form (I don't imagine doing pixel work in PS is fun, especially when things like Aseprite exist), but even those have a learning curve. It's a much shallower learning curve but you have a lower ceiling attached. The financial argument never quite convinced me; it's £120 a year for the subscription, you'd need to be several years in to justify the cost of a laptop, and I find at least that access to Adobe Fonts can be worth it too.
For non pixel work, other options exist, whether you want to talk about GIMP (whose functionality is 70% of PS's with an even crappier learning curve), or Krita, or Paint.NET or MyPaint or Corel's products (I haven't touched Corel's products in many years - I think their acquisition of Jasc made me feel uncomfortable), but even so they still all need some learning to get the best out of them. You will likely find you get somewhere faster in some of them than others, though. But no tool is going to just magically transform you or your ideas into reality.
As for using Unity in a team, I don't have a lot of experience in that department, but I can see how it could be applicable; then again I assume this is because they don't really target the team angle so much as the small indie who will be a person receiving assets from others and building the thing, and for bigger teams, Unity's source control offering exists. I do think the build server angle is a problem, though - as even their in-house team was complaining about it, and it's come up in multiple game dev conferences over the years.
In any case I hate to be that guy: most platforms have issues around multi-person development, especially in the mobile landscape. The same pain points, too: replicating environments, consistency of assets, consistency of builds, deployability of builds to shared environments. Large companies have entire teams for making this happen - have been on a couple over the years! Have also built my own build servers too, and they weren't building Unity... I don't think this is a Unity specific problem at all, but I could see an argument that Unity doesn't help itself in the way that, say, Godot goes out of its way to do so.
But I will also come back to the amount of nonsense the current game consumer talks when it comes to these games. You can build great games in any engine; you can build awful games in any engine. You could be Warner Brothers, building Arkham Knight on Unreal Engine 3 and make the PC port of the game so underperforming that you pull it from Steam. Or you could be Blizzard and build Hearthstone in Unity.
They're tools for a job, they need to be learned like any other and some do different jobs better than any others - it's not philosophical, it's just practical.