Feat: New tracing/simulation RPC endpoint

[loocapro]

Describe the feature

We observed high response latencies on eth_simulateV1, and debug_traceCallMany lacks the same override possibilities and does not return internal log traces.

Therefore, we propose a new tracing/simulation RPC endpoint that retains the override capabilities of eth_simulateV1, is as lightweight as debug_traceCallMany, and optionally returns internal log traces.

Additional context

Something like:

#[method(name = "simulate")]
async fn simulate(
    &self,
    opts: SimulatePayload,
    block_number: Option<BlockId>,
) -> RpcResult<Vec<Vec<SimulationTrace>>>;

pub struct SimulationTrace {
    /// The address of that initiated the call.
    pub from: Address,
    /// The address of the contract that was called.
    pub to: Option<Address>,
    /// How much gas was used by the call.
    pub gas_used: U256,
    /// Calldata input.
    pub input: Bytes,
    /// Output of the call, if any.
    pub output: Option<Bytes>,
    /// Error message, if any.
    pub error: Option<String>,
    /// Why this call reverted, if it reverted.
    pub revert_reason: Option<String>,
    /// Logs emitted by this call.
    pub logs: Vec<alloy_primitives::Log>,
    /// Value transferred.
    pub value: Option<U256>,
}

The difference with debug_traceCallMany would be that it computes internal logs and just expose alloy_primitives::Log and has the possibility to override state and block_env. I'd also like the idea to make the log computation optional.

What you think @mattsse?