00:38:31 re churning: given what we have already known and suspected about churning, the biggest question is whether it has a net positive/zero/negative contribution to privacy, and, if not zero, then what the magnitude is. 00:39:16 in more practical terms: due to the suspected "manual" nature of most of the churning, and maybe even regardless of that (<- IIUC the new paper implies this), it may be possible to tell ad hoc (organic) and churning transactions apart *to a certain degree*. however, it seems crucial to know what this degree is. that, together with your threat model and your available resources for 00:39:17 churning, is what would yield actionable results. 00:40:06 e.g. if the benefit is only marginal, churning may not make sense for most people, but the added "noise" may be worth it for some people with severe threat models.