...Pretty much the biggest con to waits is that only one wait can run per player at a time...
It's a good thing a bound map usually consists of one bound trigger running at a time, and the only other waits you could want would be for the players (for music and such, and death counters for that wouldn't be a bad idea, anyway).
I am a big fan of death counters.
However, speaking from experience, waits are easier for bounds. I've made roughly 20 bounds, and only 2 have used death counters for wait substitutes. The first was just an example map I made for a friend. The second was The Sims Bound, where I had 12 random obstacles running simultaneously. I've noticed that death counters don't run quite as smoothly as waits do. When an obstacle is made with death counts, you can feel it. There's something nonrhythmic about it.
For anything else, I would say death counts (they even have benefits for bounds, where the next obstacle will start immediately because there are no wait blocks). Just because death counts are more sophisticated doesn't mean they're the best solution.
On a side note, I always choose a death counter over switches for determining which obstacle a player is on. One line trumps two. If you're still using switches,
for shame.
Edit: It's also against the efficiency rule to split a bound obstacle using death counters. There's really no point in making up to 15 triggers for one obstacle when you could do it in 2. In short, it's a lot more work for little gain. Plus, adjusting timing can be a little annoying (not in the respect of changing it, just finding what to change) using death counts, where as with waits you can see several sequences at once and know where you are.
Also, not to dig under the skin of several friends of mine, but I doubt many people whose mapmaking limit is on bounds would be able grasp the death counting concept easily. It takes a little more comprehension and planning to replace waits with deaths.