I had the same experience with Apple about a year ago. Not merge sort, but an algorithm question that also involved a fair bit of math. The type where if you know the number theory trick, the implementation becomes straightforward, but if you don't, the solution is obviously suboptimal.

The recruiter also pushed me to interview with other teams, but I quickly said "no, thanks". I can't necessarily say I was frustrated though — this experience is more or less exactly what I expected from Apple or any FAANG.