I agree. As a programmer I never believe this is actual interaction of people but instead random events programmed to show up to spoof activity. There's no way to verify the truthfulness of the data. As a consequence I distrust the website and make an effort to find a different seller.
What you're trying to solve is a form of social validation and trust that brick and mortar stores implicitly have:
1. They have to have spent a reasonable amount of money to actually be there;
2. A busy store with lots of people at the registers means there's enough trust to spend money here.
To solve this in a virtual environment you'd need a comparable amount of implicit trust. For #1 it's doable: have a trustworthy domain name. Amazon.com is a lot more trustworthy than look-at-my-shop.tk.
For #2 I don't think there's a trustworthy equivalent, since it's either off-site by a third party or unverifiable by users.
These are all dark patterns used in the SaaS community and it takes zero effort to create. Two valid alternatives to this product: (1) lie (2) average out whatever proof throughput you get and simulate the events
Exactly, like the bubble/up-sell in the Uber app that claims 'busier than usual' circumstances can only be remedied by paying a little more for a quicker pickup.
I want to go one step further and say this is actually a dark pattern.