Good questions, since I still have the rocket, below is an image showing the rocket after the nose cone has been removed. Important features include:
1. The three folded/rolled streamers slightly sticking out of the top of the body tube. Slack riser-lines lead from each streamer to the top of each spar.
2. Deployed basswood spars, retained by the "taut" white string (strong upholstery thread) toward the middle of each spar, also note fishing swivels at the top of each spar. The retaining string limits/defines the maximum movement of the spar
3. The springs are at the aft end of each spar (not really shown). They are custom made from piano wire and styled after Tim Van Milligan's flexi wing glider springs.
4. Everything is attached to the unpainted body tube "sleeve" that slides over the actual BT-60 rocket body. The whole mechanism is built around this sleeve which could completely slide off the rocket's BT if needed.
5. Two of three small wire squares/loops that stick out of the nose cone. The free ends of the spars are "hooked" into these wire loops and retained close to the BT during boost, but spring free once the NC ejects.
Concerning not tangling........ believe I had the best deployment luck placing the parachute
below the streamers inside the BT. In this configuration, the ejection charge pushes out the parachute out (and the NC of course), which in turn pushes the streamers out. The spring action of the spars was
not the streamer deployment mechanism. They snap open with authority and would rip the crepe paper streamers.
Not shown are two slightly elevated launch lugs (raised about 1/8 inch) needed for the launch rod to clear all the extra "stuff" on the body tube.
This rocket has not flown in a while (made back in 2012/2013) and I don't have any before/after on-board video examples that clearly demonstrate improvement in reducing recovery rotation. Viewing the ground level videos seems to show less rotation, however. As happens often....I think I moved on to other projects, but hope this helps.