Brasas: To the gamedevs among us:
Not being that much of a techie, and certainly never having programmed at this level, I would be curious at informed opinions comparing what AMZ is making available and what Unity and Unreal have made freely available in the past.
Any obvious pros / cons of the separate engines? And I mean all around... financial, support, license limitations... not just coding.
Development-wise, seems like UE and Unity are more or less equal depending of your project, it's more a question of workflow habits, learned skills (C++ vs. Java/C#). In terms of documentation and ressources, Unity is far above the rest of engines (books, forums, videos). I don't know about UE situation but the recent tendency to implement features instead of correcting bugs has caused some anger (for example with WebGL export, they only corrected a few days ago the input from arrow buttons which was scrolling the webpage instead of being "understood" by the iframe of the game)
financial/license: The main drawback for UE are the release guidelines (mainly the 5% royalty)
https://www.unrealengine.com/release On the other side, Unity has this almost-free-for-all and "relaxed license" approach but needless to say that if you do some contract-work, clients won't appreciate to have a "Unity personal edition" as first splash-screen. Besides, the removal of the splash-screen, the Pro features aren't really useful if you're a very small team (1-3 people working on the same place). the offers on the Asset Store Level 11 aren't generally appealing (officials said on the forums it will be revamped).
Concerning, the case of Lumberyard, it's way too early business-wise to dive into it unless you can afford the time/money to learn the engine, deal with bugs, etc. But like I said earlier, a good first project to test out all the features would be some kind of Twitch-driven quizz instead of going all "let's make the next Crysis".
Now of course, things change fast in the business tech field. There are rumours that Unity 6 Pro (due for release next year) could be subscription-only (no more perpetual license then) and naturally this the kind of things where engine competition is appreciatied in order to create alternatives and not get stuck with a monopoly (like with Adobe)