I assume you are using sprite renderers or other unity 2D features for displaying sprites. If you are using custom 3d meshes for displaying 2d objects or certain materials consuming sprites as textures instead of sprites the following text doesn’t necessarily apply.
The whole point of builtin Unity Sprite atlas tool is that it makes single texture containing multiple sprites. For the most use cases it shouldn’t be inferior to manually packing or using an external sprite packer. And in many cases the workflow of being able to create a sprite atlas in unity from existing sprites without modifying existing objects or code using them is better than doing it manually. It allows you treating it like an optimization and doing it after the fact based on rest of the content. Unity Sprite atlas gives you a lot more flexibility of changing what gets packed together with what. Only after the level is done you might realize that certain partially unrelated sprites are mostly displayed together and would benefit from packing together. Using manual (or external) packing forces you to adjust your workflow around it from the beginning.
There are always exceptions where other approaches might be preferable. For example TMPro doesn’t support Unity builtin Sprite atlas and requires manually packing the sprites. In some situations good external packers might be able to pack things slightly more tightly than Unity builtin one (but it isn’t guaranteed and you need to consider whether few percent better packing outweighs cost of more complicated workflow). There is also the unity builtin skinning editor, from what I have seen it’s mostly used with sprites that come from single file (either PSB or manually packed png), not sure if it can be used with individual sprites, but that’s a more of workflow/whether feature is supported not an optimization question.
For the most part seems like your technical artist might be wrong and have incomplete understanding of how things work, or maybe you hit a specific edge case where builtin atlas isn’t functioning correctly.
For example depending on how and which resource loading strategy you are using, unity might still include and use the unpacked version, but that’s mostly an issue if you are using Resource folder. There can also be situations where even though sprite atlas packed sprites are still not batched (in which case the same reason would probably also prevent manually packed sprites from batching).
My recommendation to you is talk to your colleague . Clarify why they have the opinion they have. Did they verify that atlas isn’t working (and how they did it), are they just repeating a thing they once read without fully understanding it, or does your project has a very specific setup in which the usual knowledge doesn’t apply.
Don’t just guess whether batching is working or not. Total batch count is a useful indicator, but you can do slightly better - tools like builtin unity graphics debugger will tell for each sprite whether it’s batched or not and why. For standalone builds you can use tools like https://renderdoc.org/ and precisely check what GPU api calls the game is making, what each of them draws and and what textures are they using.
In a good work environment it shouldn’t be a problem to directly discuss the things you disagree with your teammates, even if you are in a junior position. Everyone can make mistakes and not everyone knows everything, sometimes it’s simply matter of not having spent enough time thinking about it. Most importantly explain why you think that your solution works, show the demonstrations supporting your theory like change in batch count and output of graphic debuggers. If the other person has strong ego you can try approaching from other direction, something like “Here is why I thought this way … , help me understand where I made mistake.” Which is less confrontational compared to “You are wrong!” . Either they realize their mistake while trying to explain it to you, or they explain why it’s not true in this specific situation and you learn something new. It’s a win either way. If you are very unlucky, the other person might be embarrassed to admit they made a mistake and insist on their way even if all facts disagree with it, but that’s not a healthy working environment. It doesn’t help making a better product and it doesn’t help you to grow by learning from seniors.
