The answer is, it depends on your market.
For my clients I usually suggest they write it as a website for the first version, unless there are technical limitations or the sales model is app purchase from the store.
This is because:
Once it is established you can then look at native versions because you know who wants it and what they are using. Also you can use some of the bridging technologies to like Linked In have for their iPhone app.
Another point - is the app free or paid? Apple takes a 30% cut of the pay, so that may influence your decision. Also to consider - can your application be done as an HTML5 app - bypassing all the store / platform specific issues altogether?
This is a hard question, like going to a restaurant and being told you can order food or order drinks, but not both. So my suggestion is to make an HTML app, then use something like PhoneGap to compile it into both a iphone native app and an Android native app (as well as Windows, Blackberry, etc.)
Then if performance becomes an issue, you can optimize into a native app later. (You're talking about a closed beta, so the design will change a lot from that feedback; you don't want to optimize too early.)
The quick answer is that the ROI is usually much higher on iOS apps. Startups/app developers tend to focus initially on one platform to build on. However, to develop a good app with nice looking interface, iOS provides much more, making life easier for developers.