My advice would be - if you have the money - to buy a used mac (maybe 1 year old) - that will allow you to do BOTH android and mac development.
The switch from windows to mac will take you 1 month to really get comfortable (i.e. just navigating - figuring out alternatives you were comfortable with on windows etc.).
However the development environment will be BETTER on mac - because it has BSD (unix) built into it - so no complication compared to windows where you need to stick on with bubblegum some unix stuff like cygwin etc.
You will notice that MOST google employees use mac laptops - why ?
I think for developers while windows is an excellent platform now - with the uncertainty of win 8 - and the stability of mac OS X - and it’s cross platform usability for both Android and iOS (it’s needed for iOS actually) - it makes sense to use a mac.
The reality however is that if you are doing Android - you just may not have the time to start the mac development you were thinking of doing (but since you are determined to do so - in your case switching to a mac maybe a reasonable decision).
EDIT: however, remember that on Apple App Store the requirements for apps is MORE stringent - I think they are less tolerant of R-rated apps. Meanwhile on Android it is the “wild West” where anything including copyright-infringging stuff is rampant - though technically not allowed - which results in people getting banned, then Google hard to reach - result you have this phenomenon of “I want to buy a new Google account”. On Apple you just will be rejected if your app doesn’t fit their standard. So there is that risk - perhaps one way to test the waters is to make a lite version - and see if that gets approved - then build the full version. Or if that subject already has apps on App Store - that maybe good indication that your app on same subject may be accepted.
The Android development environment on mac OS X is BETTER than on windows - no USB driver etc. needed (same as on linux I think) - the unix tools are built in (I think you need to download a free package from App Store - it is provided as a package for Xcode - so you will have gcc etc. if you need to do JNI/native development for Android).
The reasons favoring windows are if you want to do games or other stuff - but if you will be doing development 90% of the time - then you will not find any issues developing on mac for Android.
I have yet to start iOS development - though my intention was to start that - given my negative impression on profitability on Android (and it’s whole business model). Also on App Store I could sell apps - but Google Payments not available at our location.
However, over time my impression of profitabiity on Android has changed also - it is clear some types of apps will NEVER get the revenue you expect - those apps MAY be sellable on iOS (though I haven’t done that so I can’t say that with any level of authority). But other types of apps may work on Android - IF you have enough of them - though it is possible you may get a hit with 1 out of 10 apps (or your first one) - but an app that requires 6 months to develop and then fails is a risk you maybe better positioned to take if you already have 10 apps on the market and have not only some revenue buffer but also some exposure to how the market works - which may save you money when developing your longer-term game (as you may avoid doing some very simple things that you were not aware of earlier).
Also favoring Android is my perception that it will get about 5x bigger in 1-2 years - though much of that market growth will be in Asia - non-English speaking countries. The reason is cheap Android phones are inundating the market - by next year it will reach the BOTTOM tier of cell phone users. Which means the market will be HUGE.
This is one thing favoring Android - where even a non-profitable ad model may wind up making many apps profitable (which are not currently profitable). Though I wonder what an expanded market means - will ad market scale to fill the gap ? The rise of localized ad markets for the emerging markets may fill some of that gap.
So this is really militating in favor of Android - and one does then reconsider whether one should devote the mindshare to iOS …
HOWEVER, then one sees the random banning of apps and that is cautionary …
Maybe one can just start doing some small iOS stuff on the side first (loud thinking) …
If Apple wants to remain a player it BETTER start making an ULTRA-cheap version for the vast emerging markets - else it will be facing similar situation as mac vs. windows. IF it does that (and that will be great for developers also) - there will continue to be an iOS/Android alternative - which would keep Google megalomania in check hopefully to some degree (in dealing with developers).