**[WinRAR][1]**, if properly configured, can do what you described as 'Smart Extraction'. Quoting its [changelog][2]:

>Version 3.60   
New "Remove duplicate folders from extraction path" option in "Settings/Compression" dialog. 

>If this option is on and you unpack an archive which root folder has no files and only one 'somename' folder and if destination folder is new or empty and also ends with '/somename', WinRAR will exclude one 'somename' from resulting 'somename/somename' path. 

>Version 3.90   
"Remove duplicate folders from extraction path" option in "Settings/Compression" dialog is replaced by more universal "Remove redundant folders from extraction path" option. This option will eliminate redundant archive name based folders from extraction path if you unpack an archive with "Extract to DestName\" context menu command and if archive
root folder contains only one folder and no files.

![screenshot of above-described WinRAR configuration][3]

WinRAR can extract ZIP, 7z, RAR, and [other archive file formats][4]. It also supports non-ASCII, Unicode file names.

  [1]: http://www.rarlab.com/
  [2]: http://www.rarlab.com/rarnew.htm "WinRAR changelog"
  [3]: http://i.stack.imgur.com/Q3agK.png
  [4]: http://www.rarlab.com/otherfmt.htm