Same story and same reason.

How is SQL to know that the 90% free space in the log is necessary or not? Let's say the log reaches 20GB used every night at 7pm but by the time the backup runs whatever job that was has finished and the log is using maybe 50MB. If a restore of that shrunk the log to 50MB then there would be either unpleasant surprises the next night at 7pm (log growth is resource-intensive) or the DBA would have to remember to grow the log after any restore.

There's one time SQL does shrink the log by itself (reverting from a snapshot), and it's considered a huge bug/flaw.

Oh, and to do what you want, the restore would have to do huge amounts of work that it does not currently do, reading pages and allocation structures, changing page headers, updating allocation pages, index next, previous, parent and child pointers and a ton more. Would make restores significantly slower than they are now, which is really not a good thing.