Distinguishing type-in traffic from click-in traffic can be difficult. If you set up a hit counter, how does the hit counter know whether the name was typed in versus if the surfer clicked a link somewhere. . .
It can roughly be done using some Environment variables ($ENV{'HTTP_REFERER'} for you PERL/PHP junkies), but you still have the situation of anonymous surfing. When a person is anonymously surfing, the $ENV{'HTTP_REFERER'} shows up blank, so it always looks like they typed the domain into their browser. . .
You can use some indirect evidence though too. If you get 100 hits per day and you go go marketleap and check your inbound links and that number is low (say like 4) and these links are not on high-traffic sites, you can bet that you are getting some pure type-ins.
This angle has flaws too. . . For example, if I have 1000 links, but all these links are on low traffic sites, I will not get a lot of click-through traffic. However, I can have ONE link, and it is on a prominent spot on Yahoo, you bet you will get a lot of click-throughs.. So the Marketleap thing is just a guideline and not a tell all.
I do not know of any web stat program that actually differentiates accurately between type-ins and click-in traffic.
Bottom line, the number of actual type-ins (defined as if the surefer types your domain into their browser) is difficult to determine.
-Bob