In reply to tlm:
Right. In the US, we say "off belay," not "safe," which is a much better description of what is actually going on, which is that the leader has determined that he or she no longer needs to be belayed. Whether that particular leader on that particular climb at that particular moment is "safe" or not---pretending that everyone, or at least the party members, both agree on what "safe" means and are capable of determining whether the agreed-upon conditions have been met---is a completely different and possibly unanswerable question, except in retrospect.
It is true that this is just semantics, but language conditions assumptions...
Decisions that favor speed and efficiency over some arbitrary level of security are agreements that are entered into by the party members themselves, rather than concepts debatable on the internet out of the context of the climb. Whatever real safety means, it lies on a spectrum between speed and security, with the balance shifting more and more to speed as the length of the climb increases and the prevailing conditions worsen. I do think, however, that folks who don't at least practice efficient tactics on short climbs are going to be slow on longer ones because they just won't have their systems dialed, especially if the situation becomes stressful.
Personally, I've seen parties slow down in conditions that are practically screaming for more speed, resulting in unplanned bivouacs, nighttime descents, and exposure to lightning up high. Sometimes less "safety" is more safety. In this regard, Grid North and others (and I) have mentioned an approach in which the leader and second are always anchored, the concession to "safety" being that at times they are connected to a single piece they judge to be adequate. This is effective in speeding things up and is hardly a wild leap into the danger zone.
Post edited at 16:50