Koning sorts through some of the recent discussions on Bob Murphy's blog about the precedent for Fed action. I think how you react to all this depends on where your bar is for "unprecedentedness". If you think that an accomodative monetary stance in a panic is what we're talking about, then the Fed clearly didn't break decades of precedence. If you think the distinct points and policy levers of Bagehot's advice are critical for defining "precedent" then it was not following the precedent for decades, but rather the precedent of about a decade - Fed policy since the 1920-21 depression. Bob was specifically concerned about policy under the Hoover administration, and that was definitely in line with this precedent, but Koning points out that after 1932 (i.e., during the Roosevelt administration) things changed again and the scope for making open market purchases was widened.
The way I look at these things, the important point is accomodation. Policy levers and doctrines clearly change over time but that seems less important to me. The Poor Laws in the 1700s were nothing like welfare today, but the principle of a safety net was there in some form. Thomas Paine's plan was very different from social security but you still had a classical liberal case for an entitlement to social insurance. Bagehot was not Bernanke but he still discussed accomodation in the crisis. I see thsese changes less as radical departures and more as policy evolution (and we're discussing another potential policy evolution now - NGDP level targeting).
There was a lot of policy evolution in the 1920s. It was a period of major change on fiscal policy as well, with Harding's budget act and the beginning of modern budgeting practice and the establishment of the OMB. A few years later the Fed would make the major change that Koning talks about from emphasis on the discount window to OMOs.