Are we going ahead with triptych with complicated multisig in near future ? Or looking into Seraphis/lenantus3/something else?

IMO one more required multisig round (from e.g. 4 to 5) should not be a reason to disregard Triptych.

<selsta> "IMO one more required multisig..." <- 👍

There is a Seraphis-like structure that supports the current address format, with comparable scaling vs ring size (probably) to Triptych under certain assumptions. This may be the 'best of both worlds'. The benefits over Triptych would be multisig-friendliness, robust balance recovery with private view key, and delegated/delayed membership proofs (which would make tx chaining feasible).

That solution is based on in-progress research.

Under the same assumptions, Seraphis would be noticeably more efficient and would enable more flexible address scheme designs (though - not the current scheme), at the cost of large-scale implementation effort to handle a new addressing scheme.