I support the current homework policy. I "discovered" this site because some of my students hashad posted here a question I assigned in class. I was not happy to discover this but enchanted to see that the community ignored the question; it was eventually deleted.
It is not impossible to ask "good" homework-type questions and I've answered a good number myself. In the good examples, the OP makes some incorrect assumptions or makes some technical error, and it is extremely constructive to provide corrections or hints without necessarily giving the whole cake away. This is what good instructors do anyways when students come to their office to ask questions, and I'm fine with the "show me your work and I'll help you untie some of the knots" approach to answering.
However, SE-Physics could easily become overwhelmed with people simply "farming out" their assignments without discrimination. There is no value in this for anyone, except the cheaters who get their work done by someone else. Professionally, I refuse to answer questions from students unless they come prepared to show some work. This seems to be the accepted modus operandi here, and I salute this.
Can higher-level questions coexistscoexist with "homework-style questions"? To some degree they do already, and personally I find that overall the balance to be ok. The higher-level questions for me are often starting points for literature searches and fresh perspectives. The questions linked and related to individual postings are often enough so that, with very few clicks, I end up getting the inspiration I was looking for. If the balance is skewed towards too many homework questions, then it becomes a waste of time to search for useful information.