I was skeptical, so I wrote a python script to simulate old vs new PD allocation schemes. This scenario has the best possible scenario for the old (current) PD allocation system: All missiles in one volley, and all PD fired as single shots, so no overkill. In both cases, it's 100 PD against 20 missile, with a 20% chance to hit for each shot.
After running each scenario 10000 times, and averaging the results, it seems you're correct. There'll be more leakers:
Old PD scheme:
number of shots fired on average: 92.1212
number of missiles stopped on average: 18.4002
New PD scheme:
number of shots fired on average: 67.1664
number of missiles stopped on average: 13.4576
(It assumes the PD allocated to a missile stops firing once that missile is destroyed, can't remember if that's correct. Was mainly worried about missile stop rate.)
I wonder how this case looks if you actually include salvo size effects in the current (<2.2) implementation, as these are not negligible unless you are using single laser turrets for PD (which is...not optimal, anyways). The actual difference may be less than what we see here.
That said, I don't think beam PD getting weaker is a bad thing... some are complaining about beam-only fleets being less viable, but this is kind of the point behind all these missile/PD changes, isn't it? Currently, there is basically no downside to going beams-only because missiles are weaker, costlier, and more logistically demanding than beams (and Steve has shown recently that this is true even when you pursue a fighter-based doctrine, contradicting the conventional wisdom that missile bombers were the optimal form of weaponry). If buffing missiles makes going beam-only more of a challenge then I think that is a net benefit for players, at least I certainly hope no one here plays Aurora because they like how easy it is...
IMO, while the mechanics should not force this there should generally be an overall push towards combined-arms fleets as having more tactical options and flexibility should be a strong payoff for the added research and resource investment for a varied fleet composition. Yes, I can make a serviceable fleet by slapping railguns on everything and driving closer so I can hit the enemy with my sword, but this doctrine has the chief advantage of saving on weapons R&D, so maybe should not be the most effective strategy by default?
For AMM spam specifically, for beam-only fleets the solution remains the same as always, which is strategic superiority either through superior tonnage on station or superior technology (i.e. ECM). If the player is not limited to beams only, then there are more options including counter-AMM spam to distract enemy targeting. I readily agree that AMM spam is not very exciting to play against, but against the NPRs it is rarely an insurmountable problem even for beam-only fleets as it is easy enough to concentrate against NPR fleets (or spoiler... things) and NPRs do not use AMM spam as a cheese strategy with entire fleets of purely AMM ships. A player race would have that option, but then it is up to the player how they want to house-rule such a cheesy tactic, as always.