As a BAR I've only done HPR flights (first flight was my successful L1 cert), and I just successfully L2'ed earlier this month, so only H, I and J motors for me, and no flight under ~2000' or with a less than 38mm rocket/MMT. Mix of DD and motor eject also. I've built/flown 5 rockets so far. I'll consider "success" as the rocket was completely undamaged/immediately flyable, and "failure" as significant damage requiring additional parts to repair (or totally destroyed), leaving a grey-area in-between of partial damage that was easy to repair, like replacing a cracked fin fillet or cutting a small section of zippered airframe without having to replace it. By this definition, of 30 flights total, 21 were successful (70%) and 5 were failures (17%). Here's the breakdown by rocket:
1) GLR T-bolt (38mm MD, SD). 2 flights (H148R and I280DM), 50% success/failure rate. First flight was successful (had electronics issues but that only meant no flight log as deployment was by motor), second flight the chute never deployed (I believe it was due to only using half of the supplied BP in the I280DM, which was following the motor's instructions for the airframe size). So that rocket was destroyed on impact (hence the R.I.P. in my signature, take your pick between the standard abbreviation or Rocket In Pieces).
2) GLR Firestorm 54 (54mm MD, initially with DD). 15 flights (all I motors), 67% success rate, 20% failure rate (two flights had minor damage easily repaired). This has been my 'workhorse rocket', but has been through a lot along the way.
The first flight was an I215R, had an early pressure separation that led to a 1.5" zipper. So not a "success" but not a "failure" either by the above definition, rocket flew again the same day (just a bit shorter). The second flight was more catastrophic, I65W-P, first time doing DD. Drogue squib never fired (e-match went from continuity to open but never lit the pyrogen), main came out at 700' but at the speed the rocket was going zippered the lower airframe down to the supermount, and split an electronics bay bulkhead, separating the upper and lower parts of the rocket. I cut the airframe just below the supermount, installed an Aero Pack M54E (drilling my own bypass holes in it to still do motor eject) to be the new shock cord anchor, and turned the original upper airframe into the new lower airframe section above the M54E, epoxying that tube in place (using the original e-bay coupler as the tube coupler). So the rocket was converted from DD to SD as it now only separated at the nosecone and only had room for a single chute.
In this configuration (what I referred to as "mk III") it flew 10 successful flights (all motor eject) with no damage. The 11th flight (13th overall) was an I245G-L, the motor eject was several seconds later than it should have been and it resulted in a 2.5" zipper. Once again the damaged section was removed and ready to fly again. The first flight in the "mk IV" configuration was an I357T (RMS-EZ before AT started shipping with RDK parts), the pre-assembled delay assembly blew the charge 8s early and zippered the airframe all the way to the M54E. In addition at landing the Slimline threaded retainer landed on a rock and crimped itself, I had to hack-saw the retainer at its middle to get my aft closure out.
For "mk V", since the M54E anchor had the down-facing stud I gave up on bottom retention and just used the M54E and a threaded rod as the motor retainer, and once again replaced the airframe above the M54E. Since I was ordering a new tube from GLR I ordered enough (and the e-bay parts) to go back to DD. The first flight in this configuration (I225FJ, electronic eject only) I was doing an in-air test of my first flight with shear pins, so I stuffed a 36" main in both bays so that it would still land safely even if the second charge failed to shear the pins. Well the second charge wasn't the problem, the apogee charge blew but failed to pull the larger chute out of the tube. So this flight was much like the second one, first chute came out at 700' and zippered the lower airframe down to the M54E once again.
No damage to the e-bay or upper airframe this time as I took a different approach to the e-bay construction that places no stress on the bulkheads (two eye-bolts connected through a central coupler, so no load transfer via the bulkheads like with GLR's design). This is currently awaiting repair, I'm going to see if I can still get any 54mm Magnaframe to re-build this one more time for DD, now that I know the shear pins work it shouldn't have any trouble pulling the much smaller drogue out of the lower bay. If there is no more 54mm Magnaframe then I
may just go SD with a Chute Release, provided I can get it to fit along with the larger chute.
3) Binder Excel (4" DD w/54mm MMT). 3 flights (I280DM, I297SK and J350W, all SD w/motor-eject only), 67% success, 33% failure rate. First two flights were successful, third flight the motor ejection charge never fired, rocket was destroyed on impact. I had electronics on-board but no functioning e-matches, so flight data was captured even in the crashed flight.
4) RW Go Devil 38 (38mm MD, DD). 5 flights (2x H148R, I285R, H123W, I366R, electronic-eject only), 80% success rate, 0% failure rate. The only flight that caused any damage was the second one, the chutes tangled and it landed hard, cracked a fin fillet. The first flight had a pressure separation that popped the chutes early, but no damage to the fiberglass airframe. I've also had the main come out early a few times, will be shear-pinning the nose starting with my next flight.
5) Binder Excel (4" DD w/54mm MMT). 5 flights (I280DM, J360SK, J145SK-LB, I216CL, J315R, all DD with no motor-eject except for the J145SK which was SD electronic eject), 80% success, 0% failure rate. This was a rebuild of #3 (only a few internal parts re-used as all of the airframe above the fin-can was destroyed and trying to salvage the fin-can didn't seem worth it to me). Only the J145SK flight resulted in any damage, a landing with no chutes resulted in 3 fin fillets cracked, a tear in the airframe and a bit of crumpling at one end of the avionics bay (coupler). It was repaired without replacing any parts. 3 of the flights had difficulty getting the main parachute out, but 2 of those had no damage, one because the drogue-only landing was soft enough and the other because the nosecone managed to yank the chute out in just enough time before it hit the ground. For the last flight I flipped the upper airframe configuration to be a "chute cannon" instead of a "chute puller" and finally got my successful L2 cert flight.