A thought about "unlocking mechanisms"
The fundamental problem with Mystery Hunt seems to have become that we want to have an event that lasts long enough for the large power teams to not completely blast through it, but still let the smaller or less-powerful teams see most if not all of the Hunt. In recent years, we've shifted from a "solve a puzzle, get a puzzle" unlocking mechanism to one which combines "points earned" with "time elapsed"; this provides an unlocking mechanism which keeps Hunt a race for the fastest teams, but allows slower or blocked teams to keep getting fresh puzzles.
Perhaps we need to experiment with rate-limiting puzzle release.
If you have a huge team that can parallelize the research or even have multiple subgroups working on the same puzzle, and the first to finish gets to call in the answer, good for you. Your score will be better than a team that doesn't have those advantages, and that will increase your odds of finishing the runaround first. (Perhaps every team that qualifies for the final runaround gets to do it, with an ACPT-style delayed start based on your team's score.)
But you may have to sit around and wait for the next set of puzzles to be released. Bring a copy of a Thomas Snyder puzzle book or Roger Wolf's cryptics or "The Maze of Games" to pass the time. Catch a nap or grab some food to keep your physical advantage. But maybe the answer to "large teams finish Hunt too fast" is to impose a speed limit, not to tinker with the puzzles to make the Hunt baseline solve time be longer.
(I also agree with what's been said elsewhere: the Boston Marathon isn't over when the first runner reaches the Pru; the Mystery Hunt isn't over when the first team finds the coin. I think speed limits might go hand-in-hand with this approach: No one can possibly "find the coin" before 8AM Sunday; all teams see all the puzzles at 8AM Sunday no matter how far back they are, even though the weaker teams don't have enough points to be "in coin-tention"; HQ stays open until 2PM; wrapup is at 3PM.)