Parts of the download process do use zip/uncompressing. I don't know why that's not used for the first part of the download. There may be significant technical constraints. My guess? iFly runs a big batch process to build the data sets each time there's an FAA update, and in that batch process they zip things up into standard groups (all sectionals required to cover a state, all airport diagrams for a state, all approach plates for a state, etc.), but whatever data is included in the initial "aviaition databases" part of the download is a group of files that has some dependency on which set of states you select, and it's not practical to include every combination of those files in their batch process.
This is probably one of those situations that uses a reasonable set of resources to achieve a 95% good solution for 95% of users, and unfortunately you're in one of the edge cases that doesn't get the best experience.
While you wait for iFly to make improvements, you might consider workarounds that you have direct control over:
- Scoping your downloads to focus on the geography you're most interested in
- Break up your downloads into multiple sessions
- Set up the downloads to start as you're going to sleep (or otherwise to occur while you are occupied with other activities)
- Wait until you arrive in the US to download the full dataset on a lower-latency connection