Reserving judgment about the other merits of this article - it's sad to witness again how simplistic, shallow generalizations are the norm when discussing financial matters:
"Another thing which is crucial to the financial services industry is the concept of being too big to fail, which has been put to good use by Citigroup, Bear Stearns, and Goldman Sachs over the past few years in sucking money from American taxpayers."
While this may be true in some sense for Citigroup, applying that to Bear Stearns is dubious, and it's plain wrong for Goldman Sachs. Banks like Goldman and JP Morgan Chase were basically forced to take TARP investments:
http://www.businessinsider.com/uncovered-tarp-docs-reveal-ho...
Several of those banks didn't need the forced investment, didn't want it, and thus strove to repay it as soon as possible:
http://money.cnn.com/2009/03/27/news/economy/tarp_takeback/i...
Goldman in fact repaid the investment in full on June of last year, with a hefty 23% interest:
http://en.wikipedia.org/wiki/Goldman_Sachs#TARP_and_Berkshir...
In other words, Goldman didn't want or make any profit from the TARP investment - the American taxpayer did. GS didn't want that investment at all - it took it to comply with the wishes and policy of the American administration.
The current Abacus scandal is occurring not because, but despite the fact that GS collaborated with the administration and took the TARP investment. In fact, this scandal is mostly about how GS foresaw the collapse of the housing bubble, and set up positions to profit from it - thus absolving itself from any need for taxpayer assistance.
If anything, the American administration and public opinion seem more forgiving toward institutions that played the "too big to fail" card, taking too much risk and heavy loses as a consequence. Those institutions were bailed out, and are currently seeing nothing like the penalties and negative attention that GS is suffering. If the SEC's portrayal of the Abacus deal is accurate, GS should be punished - but the "too big to fail" approach is vastly more dangerous and damaging.
Either way, the discussion isn't served by over-generalizations and lumping everything together.