Update benchmarks for opcodes that somehow modify receipts

[xgreenx]

Problem overview

Opcodes: RET, BURN, MINT, SMO, LOG, LOGD, TR, TRO, RETD, RVRT, CALL; modifies receipts(usually add a new one). Each new receipt modifies BMT and slows down the addition of new receipts in the future. We need to update benchmarks to take that behavior into account.

Implementation details

Before running the benchmark for corresponding opcodes, we can fill the ReceiptsCtx with pre-defined receipts. The number of recipients should be realistic, and maybe we need to do some benchmarks to see the progression of the execution time. Without benchmarks maybe we can start with 100_000 or 1_000_000 receipts.