There are some types of questions on implementation that I think are in scope:
"Chazelle's linear-time polygon triangulation algorithm is notoriously complex. Is there any peer-reviewed research that discusses any implementation of it?"
"Fibonacci heaps are much more complex than pairing heaps, but have better theoretical bounds. Are there any comparisons of the constant factors, perhaps with experimental evidence based on real implementations?"
"Are the state-of-the-art QSAT solvers fast enough to be used in flow-sensitive program analysis?"
I think even more might be in scope if the rule is "questions are considered to be 'research-level' roughly when they can be discussed between two professors or between two graduate students working on Ph.D.'s, but not usually between a professor and a typical undergraduate student." What about questions like:
"I need to have an open-source GPU-based SAT solver as a building block for testing the usefulness of my new algorithm. Do any exist?"