Something is done now. The problem is solved.
Save for the unintended consequences of hasty action.
As an example -- one from which we're safely removed by the passage of time -- I offer the "hanging chad".
Obviously, we could not trust the results from the old Votematic machines. Something had to be done.
And in many jurisdictions, we introduced lovely new electronic voting machines with touch screens -- voting machines whose internal software was less than secure and potentially subject to tampering, with touch screens that persistently registered votes for the wrong candidate due to calibration errors, and auditing systems that would make it arguably possible to figure out just who a particular person had voted for. (The new voting machine in our district had a paper scroll that recorded the votes in order. The election judges marked down the order in which each person voted. I used it once before reverting to a paper ballot, not because I believe anyone was checking on my vote, but because I object to a system that would allow them to do so.)
And let's not forget the people who thought that Diebold, the company that made many of these machines, was going to "steal" elections. I doubt that they did, but I prefer systems that are inherently trustworthy. (Not that voting in many areas is particularly trustworthy for a variety of reasons, but let's not go there right now.)
We spent a metric boatload of money to replace a system that was defective with a system that may well have been even more defective, because something had to be done now.
There may well be occasions when something has to be done now.
There are many more occasions when it's more important to take the time to do it right and actually solve the problem that you set out to solve without breaking something else.
But usually, we do something now.