Before reading the book I hadn't realised how many more undeciphered scripts there are than deciphered ones, nor had I really thought about how there might be different levels of decipherment - it's not a binary thing. The Meroitic script (a hieroglyphic script from Sudan, related to Egyptian hieroglyphs) can be read at the level of assigning sounds/roman alphabet transcriptions of the hieroglyphs. But there are only 26 Meroitic words that have translations - all of which have been derived from the few Egyptian/Meroitic bilingual texts that exist. And it isn't even clear which family of modern languages Meroitic is related to, so hard to proceed any further on understanding it. So this script is neither deciphered to the point of understanding it, nor undeciphered to the point of being totally unintelligable (like others of the scripts discussed).
Another thing the book discusses in the introduction that I found interesting & hadn't thought about much before is the types of writing system. Obviously we use an alphabetic system, and towards the other extreme are scripts like Chinese with its syllabic system. And inbetween there are various levels of syllabic/alphabetic overlaps - like scripts that have a sign for each consonant-vowel pair. And from your undeciphered script if you have enough text then you can do stats to look at how likely the various possibilities are for this script - alphabets have the fewest distinct symbols, syllabic scripts the most. So that gives you a handle on what sort of thing you're looking for.
Luck and relationships to modern languages seem to've been key in the decipherment of those that have been deciphered - the famous Rosetta stone and other Greek/Egyptian or Coptic/Egyptian bilinguals were key for understanding Egyptian hieroglyphs. And both Linear B & Mayan glyphs wrote languages related to modern ones (Greek & modern Mayan respectively). All three also have large amounts of writing available to work with, allowing statistical analysis of sign frequency & suchlike, and also allowing potential decipherments to be checked on other texts to see if they make sense too.
An interesting & readable book, I'd definitely recommend it to anyone who's interested in languages or in ancient history.