The throughput numbers on pipelines overview page are obviously incorrect

## Expected Behavior

When I'm on the pipelines overview page the throughput numbers should be based on reality.

## Current Behavior

The throughput numbers vary randomly between -250 million and +250 million handled per second.

Sometimes momentarily I see briefly numbers between 500-2000, which make sense.

## Possible Solution

I have no idea.

## Steps to Reproduce (for bugs)

This issue probably requires a cluster setup with significant load to be noticeable.

One theory I have is that if messages come from the inputs in spikes, it impacts how the statistic gets calculated.

## Context

I was trying to figure out why the CPU usage has increased, and was checking what the pipelines are doing.

## Your Environment

7.0.3, clustered setup, EPS of 5k+.

## Checklist

[no] This issue fix need to be backported.
[no] Does this issue have **security** implications?