What does not work in Pharr? He seems to be very popular (and not just in Textkit) - which does not mean that it will work for everyone and I usually fare very badly with textbooks that everyone likes. Thus the question actually. And I am going to look up Benner's book - haven't even heard of it before.
Some people find that Pharr simply doesn't offer full enough explanations. The lessons themselves will say learn this declension, or conjugation of a particular tense and mood, or aspects of syntax merely by referencing the material, which is collected in the back part of the book, by section number. The rest of each lesson consists of Greek to English exercises, a passage from the Iliad to be translated (starting at lesson XIII) with notes, English to Greek exercises (through Lesson L II) and a vocabulary list.
If one diligently reads the indicated sections of the Grammar and skips around to find all the bits of related material lurking, sometimes, in other sections, it is possible to achieve a decent foothold on Homeric Greek.
Benner's book was written in 1903 (Google Books has an edition printed in 1910) and consists of Iliad selections, taken across the whole work, together with copious notes to the student and a brief Homeric Grammar. Unlike Pharr, however, it does not take a student from no knowledge of Greek whatsoever, it assumes at least a full year of Attic Greek.
Thanks for bringing up Cunliffe's Lexicon - it does seem like something I'll need for Homeric. Looking it up on Amazon finds the classical version and an expanded one published recently. Any ideas how the two compare (and should I get the newer one or just to stick with the standard one).
Personally, I find Cunliffe to be indispensable, mainly because one needn't worry about all of the senses and citations from post-Homeric material found in the Liddell Scott Jones dictionaries. Everything in it is strictly Homeric. The new edition seems to have added in a dictionary of proper nouns: names of people, patronymics, places. Cunliffe compiled this while working on the lexicon but decided to publish it separately to keep the page count (and cost) down. It also has a short errata list and apparently gives a fuller explanation of the system used in citations to indicate books within the Iliad and Odyssey. Otherwise, it is the same as the old edition. It only costs a few dollars more than the original edition so if you choose to purchase it, I would say get the extended version.
I have Mastronarde but don't want to start there if Homeric as a start makes more sense. And I am not sure if this is a good start for Attic - I just got it because it was there and looked interesting.
I like Mastronarde because it does give extensive explanations of grammar points (and often gives examples in English to remind students of the principles before going on to the Greek.) An answer key is also available for purchase.
In the past, most Greek textbooks assumed that students had had a couple years of Latin study and generally prepared them to read Xenophon's The Anabasis
. Mastronarde's choice of vocabulary still tends in that direction but also looks toward Plato and Herodotus.
Spiphany's points about Homer are all spot on. With Homer, there is more variability in endings and spelling in general, there are lots of words used only once, and some of the patterns of syntax are not yet as rigid as they become in Attic. However, compared to, say, Plato, Aristotle, Thucydides, Homeric syntax is much more straightforward. Also, there are lots of phrases and epithets which are used repeatedly and even whole verses that are appear more than once.
So, I would say it really is an even tradeoff between learning Attic or Homeric. It really does come down to which authors interest you the most.