fees: benchmark pallet-length-fee-adjudtment on_finalize

[gabriele-0201]

Currently, the pallet length-fee-adjustment accounts for the weight of one storage read and one write performed by on_finalize. Appropriate benchmarks should be developed to account also for the computational cost of the NextLengthMultiplier update