A nice review of galaxy classification history and techniques can be found here, or here, and you might enjoy checking out the SDSS galaxy zoo project. A good paper on modern classification algorithms (a.k.a. Bob Abraham) can be found here. Hubble's galaxy classification sequence (the tuning fork diagram) was based on optical images of bright, nearby galaxies. How representative are these galaxy types, shown below? How does viewing angle affect the elliptical sequence? How do bulge fraction, tightness of spiral arm winding, and flocculence change along the spiral sequence?

Hubble's galaxy classification scheme. [NMSU, N. Vogt]

Hubble's tuning fork, used for classification of local galaxies in the optical bandpass. What are the biases and limitations of this classification technique?

Hubble's classification of local galaxies, shown in the infrared bandpass. How does this differ from the optical picture? What sort of features and stellar populations are featured in each bandpass?