So, this is an answer to your question, and you can take the same concepts and stretch them out to Unicum DD/BB/Cruiser/whatever play.
I don't thinking reducing player capability is something that's an easy decision to make, so please understand this is academic in nature.
Recognize what the differences are, and then try to reduce the performance gap. This can also be called "Normalization" as it's bringing outliers in to a more normal center.
A Unicum/High-End player is going to have 2 main advantages: Decision-Making and Specialized Skill
First, Specialized Skill:
What is this? It's an ability to do very specific things with a high or exetreme amount of talent. This could be dodging, aiming, mechanics manipulation, and other outlier-esque situations
Flak: Some super-unicum CV drivers can avoid all, or nearly all flak. This is a Skill Check, in that a person who has the route planning/execution can mitigation most or all of this problem, whereas those with less capability in this area can not. As a way to reduce this gap, you could remove Flak and replace it with Continuous Damage that ramps simply based on distance. It removes the randomness/variety, but it brings on the players on the same conceptual field. In effect, you'd tune the losses to what an average player was experiencing and make it so the Unicum experiences the same amount of losses.
Accuracy: An average player may land half their rockets/bombs/torps, whereas a Unicum may land most or all of their ordinance. You could automate drops so that an AI takes over at a set distance to target and executes a drop the same way regardless of playerskill.
Saturation/Damage Normalization: If a Unicum is going to land 80+% of damage, and an average player 50%, then reduce the damage from hitting more then 50% of your ordinance. A 50% reduction in damage after the first half hit, or some other scaling measure. This draws damage totals more into line with standard play, but still rewards higher accuracy/execution
Specific Mechanics (ie: Slingshotting Windows): Removal of smaller, lesser known mechanics to reduce inherent knowledge-gap advantages
In general, recognize that there are a few players that will enevitably be able to perform "perfect play". Reducing the opportunities for "perfect" to yield significantly higher outcomes is required.
Second, Decision-Making Impact:
A CV's main strength is its ability to make designs spanning the entire map. Primarily, the limitation is time, but it could be argued that high-end decision making can outperform losses due to flight time.
So the question is, "How do you make the wrong choice be more impactful, and the correct choice more impactful?
Spotting: Spotting/Detection is a tool that a "Correct choice" makes matter. Spotting for the sake of it doesn't mean much, but targeted vision on specific enemies at specific times can be Very Impactful. This is a specific strength of the Class, but as decision-making is a gate here, the Normalization answer would be to reduce/remove spotting. This removes any extra reward from performing the Correct play, which would allow it to stay relevant to a less impactful decision.
There's really not much that can be done for Decision Making, as that's something outside the bounds of automation. Short of having an AI take over for a period of the match to normalize players that way and reduce player agency.