Steel actually isn't available, my map has no flux. Besides, it's sort of pointless now that I have adamantine (and lots of it) for weaponry, and in everything that adamantine isn't better for than steel, iron serves equally.
I've had to let my construction slide a bit, since I've been sort of slammed at work. I've got five more reservoirs coming up, should be done after my next normal build session. By way of making up for the silence, I'm going to post some pics that accumulated.
First off, here's my fortress plan. The grid for the network shown here would extend all of the way across the map in x-tiles, but there's still some space up top it wouldn't cover.
It's jiggered so that the longest route to any extreme from the source is 12 pipes or so, which is pretty doable based on what I've seen from the close-in fill action.
Secondly, and worth a mention for the fun factor, is that I had a goblin snatcher try to snatchinate one of my precious dwarf kids. Instead of dispatching my embarrassingly effective squad of swordsdwarves, I elected to perform a small science experiment. I imagine it played out something like this:
On a more serious, sciencey note - Magma dropped from the reservoir has some very interesting behavior. Any benefits I accrue from the pressurized pipes disappear immediately as it falls to the ground. The spread speed is very slow, as you might expect, and the form the magma pool takes after the initial settling is over is a DOME of magma, 15-20 tiles in diameter. When I say dome, I mean the magma actually supports more magma a z-level above the drop point in a fairly stable configuration. When the drop is not far along, the magma briefly stays at yet another z-level above ground. The net effect is something akin to the topography of a sunny-side-up egg, with a large flat disc surrounding a bulbous lump of magma in the middle of the drop point.
From this I conclude that there are only two effective ways to use this trap system as it stands - one, to center the enemies at the drop point. This will encase them in magma and offer the sweet spectacle of both drowning and burning at the same time. This, incidentally, is how the test went.
Two, multiple gates surrounding the enemy are opened, surrounding them in a gradually widening flood of magma that traps them in the center, eventually closing in to immolate them. This I haven't tried, but I'd anticipate pressure issues after about 20 seconds of flow. The rate of replenishment is not very high if we're talking about constant flow rate, and especially near to the source pump there's not a lot of magma in the tubes to draw on. Farther away, this might work better.
The bonus third option is one that I'm researching right now. I'm constructing a 46-pump water elevator to fill an auxiliary network of reservoirs from the brook, adding a cave-in-gun functionality to the reservoirs. The brook is rather far away from existing construction, and the elevator is a lot of work, but I've got the initial 10-waterwheel pumping station constructed and active. I sketched out a design for the water dispensing mechanism utilizing bridges to avoid having obsidian "kidney stones" (thank you, dwarves in dwarves megaproject) form and block the pipes. I'll post it once I've constructed an example. If anyone has bothered to figure out how to stop caveins from pause-recentering, I'd appreciate the procedure.