We are building distribution panels/cabinets for different customers with a supply voltages of 400V ac or 480V ac with full load currents up till 600A.
Our installation requirements manual states that customer shall execute arc flash calculations and it shall remain below <1.2cal/cm2 so no extra PPE is required for our service people and otherwise PPE is required or facility of the customer shall be adjusted (different Circuit breaker or fuses)

We actual have no real risk inside our distribution panels due to the IP3X class of the supply wiring, the Arc flash risk in our cabinet is only with the incoming supply wiring. We are also using the likelyhood No of the NFPA 70E table 130.5(c).
The fact remains that the severity is lethal which is not wanted by our EHS/safety departement.

We are now looking if we could add extra metal plating (similar as internal disconnect enclosure ) (perhaps even with ventilation holes) to reduce the severity if arc flash would occur at the incoming supply wiring.

I have seen that there some guidelines from IEEE (C37.2.1, C37.20.2, C37.20.3, C37.23, UL845, UL891) for testing but not for designing.

Does anybody have any experience or is this not feasible?

When I was working for Parker on the Solar grid-tie inverter, I was faced with a similar issue. Unless things have changed, I don't believe there is any existing data to base the construction of an arc-resistant or arc containment modification. My suggestion to Parker was to enclose the incoming power and circuit breaker in a separate enclosure mounted on the side of the energy grid tie enclosure. That way the line side of the breaker is isolated from the enclosure and only the load side (which has a much lower cal/cm^2 value) enters into the enclosure; however, it is probably unlikely you could get down to 1.2 cal/cm^2. Of course I was dealing with 1.5 to 4 MW inverters connected to the utility grid through a 4 MW isolation transformer. While some areas of the enclosure were able to be reduced to category 2 PPE, some areas of the enclosure were still category 4 PPE, just not 75 cal/cm ^2.