Darn. (Edit: does your search use only common still lives and oscillators for the fuse? I found there was a way to replace the glider with a snake that was a pretty close call, and although it would be quite rough to need to synthesize snakes or similar, it would allow a way forward)
Well, then let's get serious to find out how hard this will be.
For SE rakes, the important factors are the phase, the parity of the x displacement and the value of the y displacement. The exact x value isn't important because the climber's position along the track can be freely adjusted in 4 generation increments, which results in a 2-cell horizontal offset between SE rake positions.
Now, between gliders on the same track there is a displacement (19, 45) = (72/4 + 1, 72/4 + 27), and no relative phase. The climber causes a track displacement of (4, 20), and also no relative phase. This means our climber alone affords us no control over x parity, over phase, or over y value mod 5. (Regarding x parity: even though the x separation between climbers is 19, the x separation between stationary targets is 1 and between SE targets is 1 - 72/4 = -17, so the parity remains out of our control)
That's pretty bad. Where in the waterbear we had (23*8 + 79*2) = 342 tracks and gcd(342, 95) = 19 cosets under the climber generator, here we have (27*8 + 72*2) = 360 tracks and gcd(360, 20*8) = 40 cosets. Without a different way to alter the track, we have 40 distinct types of SE rake interactions that any given track pair will only ever be able to make a single one of.
Okay, but in the waterbear it was natural to talk about SE rake interactions because SE gliders were everywhere and our real only source of material. Here we have a massive spark to directly connect streams to generate our gliders and targets, so the above might not be so bad. But it is still inevitable that we will need to use SE or NE gliders to target stationary debris for syntheses.
What about NE gliders? Since our base reaction travels on SW gliders our constraint here is the lane. With the waterbear we had 23-5 = 18 lanes, and our climber generator moved the track by 3 lanes. That left us 3 cosets, which was inconvenient but workable. Switching cosets was the reason for some of the beehive->beehive slow transformations. Here, we have 27-1 = 26 lanes, and the base reaction moves 16 lanes. 2 cosets. Okay, this isn't perfect but better than the waterbear. Unfortunately, we don't have frozen tracks in the form of beehives so we can't reuse the solution from before.
Basically my current assessment is
pros:
sparky reaction with at least 5 pairing options
relatively cheap helix
cons:
no way to reuse the waterbear's solutions to the too-many-cosets problem..
Edit2: on a happier note, there is a simple blinker -> NE glider reaction.
Code: Select all
x = 88, y = 200, rule = B3/S23
2$71bo$70bo$70b3o3$85bo$83b2o$84b2o38$52bo$51bo$51b3o3$66bo$64b2o$65b
2o20$37b3o18$33bo$32bo$32b3o3$47bo$45b2o$46b2o2$36b3o27$35b3o9$14bo$
13bo$13b3o3$28bo$26b2o$27b2o11$34b3o5$4bo$3bobo12bobo$2bo3bo11bo2bo$3b
o2bo10bo3b2o$3bo2bo10bo$4bobo$17bo2b3o$17b2o$19b3o$19b3o5$3b2o$3b2o$
10bo$9bob2o5b2o$6b3o11bo$5b5o8b2o$4bo2bo2b3o2bo3b2o$4bobo4bo2bobo2bo9b
o$5bo3b2o3bobobo9b3o2b3o$10bo4b2obob2o5b5o$7b3o12bo4b3ob2o$7b2o9bob4o
4bo2bo$22bobo3b3o$23bo2bo$24b2o6$35b2o$36b2o$35bo!
Physics: sophistication from simplicity.