First request always takes at least 200 milliseconds

### Expected Behavior

The first request should match the latency of the subsequent requests. The subsequent requests are in the impressive 0.7 - 1.0 ms range, which is faster than Postman. By quickly clicking "Send" over and over, the subsequent requests get this time.

### Actual Behavior

The first request is actually 200+ ms. Postman is not inconsistent like this.

### Reproduction Steps

The "first" request happens if it's the first request after opening Insomnia, or by waiting 5 seconds before sending it.

Sample code:
```js
const http = require('http');
const PORT = 3000;

http.createServer(function (req, res) {
res.writeHead(200, {'Content-Type': 'text/plain'});
res.end('Hello World!');
}).listen(PORT);
```

Prerequisite for my sample code:
- Install [node.js](https://nodejs.org/en/download/)
- Pick an available port for the script.
- Start with `node sampleScriptName.js`
- Open Insomnia and add a GET request `http://localhost:3000`

Method 1:
1. Open Insomnia
2. Send the request
3. It takes 200+ ms

Method 2:
1. Insomnia is already open
2. Wait at least 5 seconds after the previous request
3. Send the request
4. It takes 200+ ms

### Is there an existing issue for this?

- [X] I have searched the [issue tracker](https://www.github.com/Kong/insomnia/issues) for this problem.

### Additional Information

I have also reproduced the error in Insomnia 2022.5.0-beta.5

### Insomnia Version

2022.4.2

### What operating system are you using?

Windows

### Operating System Version

Windows 10 Version 21H2 19044.1826

### Installation method

download from insomnia.rest

### Last Known Working Insomnia version

N/A

Reproduce the delay using the provided Node.js HTTP server and an Insomnia GET request to localhost, comparing the first request with requests sent within five seconds. Investigate the request lifecycle around opening Insomnia and the idle period; done means the first and delayed requests have latency comparable to subsequent requests.

Written by the indexing model from the issue text.