I'm of the school where if you want it to separate, hit it like you mean it! If it works on the ground nicely and smoothly, it might not work in the air when the rocket may not be stationary, may be vibrating, just hit temperature changes, air pressure changes, got a little debris in the joint, absorbed some moisture in the joint... I don't have a problem generally speaking of using one oversize charge, with recovery system designed to handle that charge, and multiple altimeters with ematches into that one hefty charge.
Be aware that a large charge for drogue deployment may generate enough shock to separate the main. Same can be the case with a larger than necessary drogue, or deployment of drogue with notable velocity. So size shear pins accordingly, and work on your harness so it progressively absorbs the momentum of the sections rather than slamming to the limit - and shearing the other shear pins or slamming all the segments together and then possibly tangling.
Most people worry most about getting the main out. Ballistic with a heavy rocket and then getting the main out tends to rain segments down at high speeds as it rips the recovery harness apart - which is generally better than coming in at high speeds all in one piece. But I'm more concerned with getting the drogue out. If the main doesn't deploy, the speed is at least managed by the drogue and you get to see it coming down and hopefully not be under it. If one of the deployments fails, it can be safer if that failure is the main not the drogue. This can depend on the details of course.
If you are going to 1000' and deploy main at 700', make sure to get the main out. Go to 15000' and deploy main at 700', be sure to get the drogue out.
BTW, I'm talking about BDRs. And of course you want to get both deployments to work and design so they won't tangle so it doesn't become a lottery how it comes down. You don't want that winning ticket.
Gerald