Calling surveys "math" is just another failed attempt at distracting attention from the lie you told in your OP
Now that I have a little more time, here is the math:
Gallup shows uninsured rate falling from 18% (I'm taking the peak, even though that's not really fair) to 13.4%. Each 1% represents approximately 3 million people. The MOE is 1% - so, just for ease let's change it to 14% (could be a little worse could be a little better, but since I took the peak, it seems fair). That represents 12 million previously uninsured gaining insurance.
From the 12 million we estimate picked up insurance... Some 8 million obtained it through employment - which agrees with what most economists have been saying (not necessarily the number, but that most picked up through employment) I've previously posted links) and can be seen in the Gallup chart showing a steady increase of uninsured from around the start of the "great" recession in 2009 as people lost jobs....
That leaves roughly 4 million previously uninsured picking up insurance through the exchange and Medicaid....
let's say the entire 4 million gained insurance through the exchange.. That is only 50% of the 8.x million that signed up through the exchange.
We know some portion of that 4 million actually gained insurance through Medicaid - whether due to people that have always been eligible but didn't sign up or due to the expansion.
So, it would certainly be less than 50% of those that signed up through the exchange were previously uninsured.
Even if you want to quibble with some of the numbers, you would still end up with low numbers of individuals gaining coverage through the exchanges.
I realize, it's all back of the napkin type stuff.. and I realize I can't say for certain, but we have a hella good idea of the numbers, which if you recall, is what I indicated above.