as it says on the tin; each laser selects a piece to fire on, cycling through all 'valid' pieces until it finds the piece with the lowest number of turrets already on it.
this could be accomplished by keeping a variable attached to the pieces of salvage counting the number of salvagers actively working on it, salvage laser checks pieces in range/line of sight, and cycles through all the list, then verifies the piece it first saw with the lowest number of salvagers already on it, and if it hasn't gained any new salvagers, that's it's target.
this allows for fighters, as well as multiple ships, to salvage entire fields of salvage without getting in each others way, stripping the smaller pieces first, which means less debris left over, less missed resources, and faster overall salvage, as the larger pieces will stick around until the small pieces are removed and turrets/fighters can focus their DPS on the remaining heavier pieces.
splitting DPS may seem counter-intuitive for faster salvaging, but consider what happens when you have a ship with 10-20 salvage turrets on it (dedicated salvager ship).. the entire selection of turrets selects a single piece, and holds fire while the turret adjusts to target that piece. the first couple of salvagers align and start firing, but the piece was a tiny one, and it's destroyed. ALL the salvage turrets stop firing, select a new target, and start turning... meaning that some of those turrets might NEVER fire on a single piece of salvage in the entire 'operation'.. it's wasted DPS.
now consider how it would behave with split-DPS.. each turret would only start focusing on a piece of salvage that isn't already being targeted, drastically increasing the probability of the piece still being there when the turret finally turns to fire on it.
consider the implications of using this logic with fighters, they frequently stop salvaging to 'reposition' then begin a salvaging run moving closer and closer anyway, if each fighter is assigned a separate piece, that piece will likely be there throughout. as the number of valid targets diminishes, likely because they were the smaller, lower HP pieces, the turrets/fighters naturally begin to focus fire on the remaining pieces, bringing the single-target-DPS back up where you'd expect, resulting in those massive central pieces being destroyed last, but much less potentially valuable resources/loot being left behind in the end.
especially coupled with 'skipping' pieces that are the 'parent' of another piece, you end up reducing the frequency of wrecks being broken into chunks lower than the AI threshold for salvaging 'worthiness', resulting in even less wasted resources/loot, while also reducing the frequency of needing to re-root wrecks and copy the new wreck into a separate entity