I'd like to make the case that speeding, while illegal and dangerous, is incredibly beneficial when done over long distances, assuming one values time, and that conversely, when done over short distances, is relatively useless. 

Let's say the distance between my house and my job is 10 miles.  And let's say the speed limit is 40 mph.  If I drive the speed limit the whole time, my commute will take
(10 miles) x (1 hour/40 miles) x (60 minutes/1 hour) = 15 minutes
Let's say I go 10 over the speed limit, i.e. 50 instead of 40.  That means it'll take 12 minutes to get to work.  That 25% increase in speed equates to a 20% decrease in time.  Not bad, but we're talking pennies here.  Three minutes?  The $120 speeding ticket doesn't even come close to making it worth it.  That's $40 per minute! 

But what if, instead of 10 miles, I was driving 100 miles.  And again, let's say the speed limit is 40 but I go 50.  What would've taken me 150 minutes (2.5 hours) now only takes me 120 minutes (2 hours).  It's still a 25% increase in speed and a 20% decrease in time, but suddenly the time is starting to add up.  Thirty minutes?  I can do a lot with an extra 30 minutes. 

Now let's say instead of 10 or 100 miles, I'm driving to Orlando, FL, which is about 1000 miles away.  Beep boop beep, multiply by ten and suddenly I've shaved 5 hours off a 25 hour trip (obviously the speed limits are higher, but still). 

What we have here is a basic lesson in percentages, specifically the idea that a percentage of a larger number is greater than the same percentage of a smaller number.  It's simple, but it says a lot.  It says that speeding over short distances is hardly worth the time savings or the potential monetary penalty.  But it also says that by speeding over long distances, you can save huge chunks of time, and that regardless of whether you place a higher value on time or money, that $120 speeding ticket could actually be worth 5 extra hours, which is only $24 per hour.  You can't argue with math. #travel