Honor's Second Marsh strategy
Okay. The key point in Honor's strategy at Marsh pitvoted on the fact that she was virtually certain she was going to be attacked and that the enemy didn't know what strength she had. The real reasons she tucked the [Protector's Own] away in hyper were:
(1) To carry out precisely the tactic she finally employed, if possible;
(2) To make certain that if Tourville managed to sneak a stealthed recon drone through the system (which is not impossible, especially in a system -- like Marsh -- which does not have the huge passive arrays which would make a quiet hyper translation impossible to hide) he would still not know about the PO until she chose to let him learn about it.
The ambush mechanics itself were not difficult. First, remember that sensor conditions in hyper-space are routinely far worse than they are in n-space. This is one reason it's so much easier to lie hidden there in order to ambush something like a convoy. What Honor did was to deploy the PO just outside the hyper limit of the system on what she expected would be the Havenites' most likely approach vector. Then they shut down their drives and went to stealth. When Tourville & Co. arrived, the hidden PO was able to track them without being detected in turn, although the fact that the PO itself saw him wasn't really particularly crucial to Honor's plan. Once Tourville dropped back into n-space, Honor was able to track him real-time with her FTL scanners and to project his approach track. At that point, it was simply a matter of timing the moment at which her deployed DD (again, beyond the hyper-limit and contactable in real-time only because of the FTL com) translated up into the alpha bands and delivered her order to the PO to bring up its drives, make its translation, and come in behind the enemy at the time and exact position she provided in her message. Even if the Havenites had approached from the opposite side of the system, she still would have had plenty of time for the PO to run around to the other side of the system in hyper to come in behind them, although the geometry might not have been quite so favorable for her.
Please note that it is largely the existence of the FTL sensor net and com which made this tactic possible. For quite some time, many people have been noodling about the possibility of using a "defenders tucked away in hyper" ambush scenario as a system defense tactic, and I have been pointing out that it would be impossible for the defenders to coordinate well enough. Well, the FTL com improves coordination and detection ranges considerably, thus making the tactic practical. It still relies on knowing -- or at least very strongly suspecting -- that an attack is coming in within a fairly specific time window, since there are practical limits to how long you can hold a fleet in stealth. But I would suspect (although, of course, I don't know for certain) that this tactic may be used again. And, at the very least, attacking admirals are going to be looking over their shoulders and worrying that it might be used against them.
IOTW, this is not the sort of "one-off, single-time surprise" tactic she got away with at Cerberus only because no one expected her to be lunatic enough to try it.