Whatever the encryption or other stake raising argument is found today
or in the future, the point is, deviation from line oriented interfaces
for the installer is not the way it can be handled by other systems,
meaning it's not the least common denominator that lends itself to
machine processing and there is point in improving this but going in
the reverse direction counter complexity and contrary to the topic
statement, and towards simplification of the interface like controlled
keyword subsets (don't confuse this with a domain specific language,
think partitioning templates), though it does not mean it can not have
more than one front end, yet modifying the installation process to
accommodate use of multiple different types of installers is not
efficient so far in terms of lack of resources.
