And you can get a macbook air for around $900-$1300. That's all you really need, nothing fancy.
And if your app is meant to be mobile, you can get deep into mobile web development and even create a very native experience. That's what forecast.io did.
Also as you probably know, in startups it's more about proving a model than having a fancy piece of software. Think about what features are going to attract customers asap. Iteration is easier when you have customers giving you feedback.
IMO RoR is the fastest way to make those vital iterations. Once you're sitting on your $5 million in venture cash you can worry about architecture and scalability.