Actually Sarbanes-Oxley was passed because big corporations that were decades past IPO defrauded people. Frauds at IPO were already covered under other legislation and were rare and easily prosecuted. It was WorldCom and Enron (among others) that lead to this legislation not IPO scams.
Arguably you could achieve the same results (auditor independence, senior management accountability) without any change to how small public companies behaved. Enron and Worldcom could bend auditors to their will because of their size, <insert tiny company worth under 10mil here> already could not do that. Why not have the requirements scale with size then?
Personally I think Audits will continue to be worthless until Auditors are required to have some "skin in the game": if an auditor signs off a company's accounts and it turns out there was a mass fraud or bankrupt, the auditor is civilly liable for at least some percentage of the loss. That's no different to how rating agencies should also have been reformed. "If you want to know how risky something is, don't ask a 'consultant' paid by the guy selling it, get an insurance quote".
As it is, these things are always trade offs: is it better to lose $1bn in fraud or spend $1.5bn in extra accounting and investigation? If new rules remove 2 frauds but also prevent 200 companies going public is that better or worse? If the US brings in rules and companies just list on other markets and the same investors get defrauded, does any of this matter? And how does anyone even measure the "what if" scenarios with any rigor to inform our decisions? It's not (IMHO) a solveable problem...