bottomline: if you are working offline with selects on large chunks of data, MyISAM will probably give you better (much better) speeds.
there are some situations when MyISAM is infinitely more efficient than InnoDB: when manipulating large data dumps offline (because of table lock).
example: I was converting a csv file (15M records) from NOAA which uses VARCHAR fields as keys. InnoDB was taking forever, even with large chunks of memory available.
this an example of the csv (first and third fields are keys).
USC00178998,20130101,TMAX,-22,,,7,0700
USC00178998,20130101,TMIN,-117,,,7,0700
USC00178998,20130101,TOBS,-28,,,7,0700
USC00178998,20130101,PRCP,0,T,,7,0700
USC00178998,20130101,SNOW,0,T,,7,
since what i need to do is run a batch offline update of observed weather phenomena, i use MyISAM table for receiving data and run JOINS on the keys so that i can clean the incoming file and replace VARCHAR fields with INT keys (which are related to external tables where the original VARCHAR values are stored).