Faulty numbers on Obamacare signups

It looks as though the numbers touted for Obamacare signups are completely misleading, even disregarding the blatant efforts to massage them.

An article by Sean Trende has come and gone, but it is hugely important. He argues that the supposed 4 million signups for Medicaid in October and November are not net new additions attributable to Obamacare. Trende's critical point is that the Obamacare proponents assume every Medicaid signup was due to ACA, but that is ridiculous. How did Medicaid get to 60 million members before ACA? How many sign up every month on average? Why are signups in states that did not expand Medicaid about equal to those in states that did?

His analysis suggest that maybe 200,000 of the 4 million signups can be attributed to the law.  Evidence: half the signups are in states that did not agree to Medicaid expansion. Medicaid, with 60 million already enrolled, has new signups (and also loses some enrollees) every month as is.  Ezra Klein and company have been arguing that 4 million Medicaid signups in two months (likely 6 million through December), plus 2 million in exchanges, means total signups are 8 million, above White House estimate of 7 million total in first 3 months.

But if in fact Medicaid enrollment increases attributable to Obamacare is 100,000 a month, or 300,000 through December, and the two million signups on exchanges include many who lost insurance in the individual market, then net due to the ACA including Medicaid in first 3 months may be trivial, and  matched or exceeded by those who lost coverage and did not get policies renewed, nor sign up on exchanges. The key number is not how many of the 2 million exchange signups (state and federal  combined) are young, but how many were simply moved over from individual policies they had cancelled.