The advantage of using accessors is that your implementation can switch between static properties and computed properties and the consumers of your API don't ha

The advantage of using accessors is that your implementation can switch between static properties and computed properties and the consumers of your API don't have to know which is which. This is called the uniform access principle and ends up being extremely useful. For example, imagine you have an Store object that saves the tax rate: Store = SC.Object.create({ taxRate: 0.0895 }); Since this is a static value, I don't technically have to use an accessor. But imagine my state implements a tax holiday, and so for one day per year, the tax rate is different. In Amber, this is a simple change: taxRate: function() { if (this.get('isTaxHoliday')) { return 0; } return 0.0895; }.property() Instead of trawling my codebase to references for this property and changing them to method invocations, because everything goes through get(), I know the change will be transparent. It's this transparency that leads to very maintainable web apps. Some people object to the extra method invocation, but I think the expressiveness it gets you is worth it. As soon as proxies land in JavaScript, we'll add support for them so you can use the dot notation and still get the same functionality.