The situation shouldn't change much as it is today. Probably AT&T reception issues would be almost ironed out, and AT&T will get more phones in, so more choices for customers.
But until the merge actually happens, which is more than a whole 12 months away, you won't likely see any immediate effect. And it's unclear whether AT&T will leave T-Mobile as a separate entity or absorb it completely.
If AT&T leaves T-Mobile as a separate entity, which is totally understandable given that absorption would be quite a financial and restructure nightmare, then nothing would change much. Maybe T-Mobile and AT&T will overlap less in their services and rate plans.
If AT&T absorbs T-Mobile completely, which means T-Mobile all over the world (mostly in Europe) will be AT&T, then that's implication that the next iPhone on AT&T will be an unlocked or world phone version that can be used on any GSM network. That along with boosted reception and service performance will give Verizon quite a run for its money. For Android, it'll mean quite a competition as the iPhone's biggest obstacle right now is its carrier-locked nature in a number of markets. But aside from that, it shouldn't have any impact on Android.
What would have an impact on Android would be if Apple suddenly comes out and drops the iPhone price down, or introduce a new phone at a lower price point. But that's for another discussion. The AT&T and T-Mobile shouldn't have any large impact on any mobile OS aside from implication that there may be hardware changes abound for future phones to cope with both networks at the same time.
Of course not. Android is #1 in the world, it'd be dumb for carriers not to support it. They've always had high end and low-end.
Whether Android is #1 in the world is highly debatable since we have no concrete worldwide data to base any conclusion on.
Saying Android has the most marketshare in mobile OS in the U.S. is more likely to be true.