I'm not sure that's much of a secret...
Well, it seems to be, since we apparently fear Castles. It's only worth fearing Castles if we can't hit hard (Bombard with multiple units) before the Castles get built.
I've seen a lot of misunderstanding about Bombardment on the forums and I have yet to see it explained properly until now, so I think that there is a lot of misunderstanding on the subject out there.
Anyway... what we cannot control is Cathy choosing to manually build a Castle. If she does so, then we are equally going to suffer whether we declare war on T + 7 or T + 11.
If she does not manually build one, then we should be landing a lot of Trebs (4 to 5--and I'd prefer 5 at her capital, since it not going to have any chance of having only 40% Cultural Defences with 2 Wonders in it) on the first turn of war, so that we can Bombard away the defences before a whipped Castle will have any real effect.
If we can do so, then we can actually LAUGH at Cathy if she whips a Castle, as it will be of less use to her than a Longbowman--IF we have a good amount of Trebs (the more Accuracy-promoted ones, the better, in my books) that land on the first turn of war and Bombard on the second turn before she will be able to whip a Castle. Even if she has already manually-built a Castle, having more Trebs will give her less time to build a ton of Longbowmen.
I really, really want to hit two Cities on the first turn of war. However, being practical, if we're going to do the northern passage, I'd rather just hit Moscow with 4 Galleons' worth of troops (using 5 Trebs + 6 Maces) and not hit a second City at the start of the war.
There is always the chance that non-threatened Cities will not whip Walls/Castles/as many units, so I would rather take that chance than attack 2 Cities with 2 or 3 Trebs each and then find Castles going up in both Cities due to the proximity of our armies at both locations but us not having enough Trebs to Bombard down the defences in time to make the Castles relatively ineffective.
So, if we vote for the northern passage, then I am going to push VERY HARD for both:
1. Only attacking Moscow on the first turn of war
and
2. Using 4 Galleons to do it
I still prefer delaying the war slightly (T + 11 instead of T + 7), since I think that "getting the war started sooner" is not equivalent to "ending the war sooner" when it comes to fighting an AI with Feudalism + Engineering. I actually think that the war will end more successfully (sooner and with less casaulties) if we strike with two reasonably-sized forces of Trebs backed by Macemen, and ideally I'd like to hit with 5 Trebs as Moscow and 4 Trebs at St. Petersburg, using 4 Galleons at Moscow and 3 Galleons at St. Petersburg.
We have enough Macemen already (6 Maces to Moscow and 5 Maces to St. Petersburg) as long as we don't lose any in our fighting, but we're short on Trebs (I'm hoping for 5 + 4 = 9 and we have 7 right now).
We can create 3 units from Willem's Cities by Turn 10, which allows us to lose 1 unit. Unfortunately, waiting for the 2 units created on Turn 10 to get into Galleons might take a while... okay, I'll open up the game... we could get the unit from Amsterdam into the war (i.e. landing next to Moscow) on T + 11 but we couldn't quite get the one from Rotterdam there until T + 12 (landing next to St. Petersburg). Thus, I don't think that we can count on that unit from Rotterdam.
So, if we take zero battle losses between now and then, we can strike on T + 11 with my ideal force of:
5 Trebs + 6 Maces at Moscow
4 Trebs + 5 Maces at St. Petersburg
and we'd just forget about that last unit coming out of Rotterdam.
Taking zero battle losses is a tall order, so obviously, if we can ship in a couple more land troops from our core Cities, we'll have some breathing room, without having to wait until after T + 11 to launch the attack.
We're really talking about the difference of 4 turns: T + 7 versus T + 11, but those 4 more turns will allow us to better prepare our forces to have overwhelming numbers.
Honestly, let's think about it. If you want 4 Galleons to sail out to another front (to Vicky or wherever), will we have all 4 Galleons ready at T + 19? If the answer is "no," then why can't we use the 7th Galleon against Cathy for the first turn of war? One Galleon can then sail back to the east, since after taking the first two Cities of Cathy's, we'll be facing Cathy's remaining Cities that:
a) Have less population points to whip
AND
b) Have less infrastructure (less likely to have a Forge)
AND
c) Would first need to build a Wall before they could build a Castle, so there's a bit of time there
The 7th Galleon that we own now can become the "last Galleon" to take part in the second front... unless you think that, say, T + 19 would be too late for it to arrive in our Core Cities' area, then it really is free to help out in dropping off troops for the first turn of the Cathy war. That "T + 19" is an estimate because I'm not sure which AI we're attacking, but that number would be smaller if we were attacking Joao or Isabella as part of our second front.
So, what do we prefer: Starting a war 3 to 4 turns sooner (which we may or may not have enough units for) or flexibility to use said 3 to 4 turns to capture an additional city or two, whip more units, etc.
While it could be less than 4 turns worth of difference to attack using the "Fort the Pig" approach, we might as well hit in force if we're going to delay these extra couple of turns, which to me means hitting on T + 11, even if it possible for the Fort to allow for a T + 10 attack. So, let's just think of it as a 4 turn difference.
To me, it's all about dropping a lot of Trebs off at each of Cathy's Cities (or at least at her 2 Walled Cities) at a time.
So, if we hit 4 turns sooner, I say send 4 Galleons and just take Moscow with a large force and forget about attacking St. Petersburg for a while, even though doing so means having boats out of position and not much of a chance to get a Fort anytime soon, but at least then we can focus on attacking Cathy's northern Cities, while hoping that not having units near her other Cities will keep her from spamming too many Walls + Castles + units.