There are two main concerns here:
Migrating questions to beta sites is strongly discouraged. This is mostly because beta sites are still in testing and there is no guarantee that they will succeed. Many of them do, but some (like Theoretical Physics) don't really generate enough traffic, questions, experts, and overall liveliness to make a sustainable critical mass. Migrating questions to a site that then gets closed is a poor service to the OP. For some sites, like Chemistry, this was not really an issue (and indeed Chemistry just graduated to a full site), but Engineering is very new (about four months), so migrating content over there should be done on a case-by-case, as-needed basis.
An additional reason for this is that if the migration channel gets too big, even a small fraction of the questions from a big site can be overwhelming for a small beta site. We want the community behind Engineering to come up with a site flavour of their own, not to build a site around Physics exports.
More generally, though, default migration channels are generally quite tightly controlled, because flaggers will know what's off-topic here, but they don't necessarily know what is or is not on-topic on the target site. Within the SE network this is a bigger issue with large sites like Stack Overflow and Super User, but as a result the SE team is very reluctant to add default migration pathways to any site.
However, this does not mean it's impossible to migrate content. Moderators can migrate to any site and will do so if you flag the question and the migration is appropriate. There is at present a low load of migration candidates and it can be managed well by the moderators. Within the past 90 days there have been 39 migrations to math, 19 to electronics (with 15% rejected), 9 to History of Science and Maths (with one rejected), 8 to Chemistry (with 2 rejected), and 5 to Engineering. It's really not worth it to set up a default channel - even for Chemistry, which is already graduated.