I’m sure someone has mentioned this here, but I haven’t taken the time to read all the comments so I apologize if I’m repeating. The difference you’re not noticing happens when a pixel “alias” and looks blurry – which you can see on the buttons of the 1x version you created. Since the 2x is twice the screen size of the 1x, you must always design in numbers divisible by 2, ie; your button stroke should be 3px because when you convert to 1x it becomes 1.5px and you can’t have .5 pixel… it’s blurry and for all intensive purposes, wrong or bad design. You don’t particularly notice it the way you notice the color being different, however there is always a feeling that something about the design isn’t right. Once you snap those vectors of, say the buttons to the pixel grid and it looks sharp and crisp, it’s the ah-ha moment. Most people won’t notice it outright, however, they do notice the difference between a good design and a bad, and most of the good designs adhere to the pixel-snapping principal. Imagine an architect not taking the time to line up and exact his blueprints, the building might stand up but it won’t be perfect or safe – and who wants that.