Should the time spent on developing a new software product be proportional to the market share of the platform? For example, if Android is 50% of the market share and Apple is 25%, then for every 2 hours spent developing an Android app, you should only spend 1 hour on iOS.

Is this the best best strategy money-wise for a startup that needs to develop similar software for multiple platforms? In which cases would you go against this seemingly common sense strategy?
Another tricky place to allocate resources is in website development. There are still large chunks of people using Internet Explorer 6, 7, and 8. Even though IE6 has the lowest market share, it's the most difficult to program for. So allocating less time for IE6 doesn't always produce the best website in IE6. Although, others argue that since IE6's market share is so low, it doesn't matter if you don't have time to put all the fancy features in it.
