If you wanted to keep a second "burner" phone. Your own prior generation phone would be substantially less suspicious option compared to buying a second device. iPhone users may well in fact prefer an iPhone either a relatively cheap model if they are of the affluent set or a used model purchased for cash. There is absolutely no reason to believe that this would eliminate all burner phones.
Lets do some speculation to firm up the idea. 55% of US users have an android, 45 an iphone and our overall probability of talking to someone who already has a girlfriend is on the overall pretty high say 50% with half of those having a second phone and 80% of those being cheap androids.
The overall percentages of users using each platform is actual the rest is speculation. I have actually probably WAY overestimated the numbers to prove that the strongest argument is basically still bad.
Take 1000 suitors. 500 are cheating bastards. Your chance are 50/50. You apply your first pass green bubble filter and filter out 550 including many good matches.
Lets look at how this effects the cheater pop. We start with 500. 250 aren't using a burner phone and 113 passed the green bubble test another 50 passed because their burner is an iphone.
Your first pass filter only reduced the cheating bastard percentage from 50% to 36%.
Now if we reduce the cheating bastard percentage to a hopefully more realistic 25% we are talking about it reducing the percentage from 25->18.
The worst thing about this algorithm is that it makes it impossible to bubble up especially good choices who nonetheless fail the green bubble test. It's also one that potential bad choices can trivially game by buying a 4 year old iphone on facebook for $90. It's the worst of all possible worlds.
> For every complex problem there is an answer that is clear, simple, and wrong.
H. L. Mencken
Talk to people like human beings and you should be able to more reliably pick out the cheating bastards.