I will briefly discuss some of the main techniques involved in showing time lower bounds for streaming problems, focusing on one in particular, the information transfer method. I will then go on to present some of the barriers to making further progress, presenting a simple looking mathematical conjecture with a probabilistic combinatorics flavour that derives from this work. The conjecture is related to the classic "coin weighing with a spring scale" puzzle but has so far resisted our best efforts at resolution.