GAE deployment (standard F1) outperforming (standard F2) with identical source? - google-app-engine

Question

I have 2 GCP projects "dev", and "prod" for a simple NextJS app. I am deploying to App Engine with the following configurations:
app.dev.yaml
runtime: nodejs16
instance_class: F1
handlers:
- url: /.*
secure: always
script: auto
- url: /icons
static_dir: /public/icons
app.prod.yaml
runtime: nodejs16
instance_class: F2
handlers:
- url: /.*
secure: always
script: auto
- url: /icons
static_dir: /public/icons
next.config.js
const nextConfig = {
experimental: {
externalDir: true
},
reactStrictMode: true,
images: {
domains: ["ik.imagekit.io"]
},
distDir: "build"
};
module.exports = nextConfig;
package.json (scripts)
"scripts": {
"dev": "next dev",
"build": "next build",
"start": "next start -p 8080",
"deploy:dev": "npm run build && gcloud app deploy --quiet --project=<dev-project> --appyaml=app.dev.yaml",
"deploy:prod": "npm run build && gcloud app deploy --quiet --project=<prod-project> --appyaml=app.prod.yaml",
"lint": "next lint --fix",
"format": "prettier --write './**/*.{js,jsx,ts,tsx,css,md,json}' --config ./.prettierrc"
},
My dev environment (F1 instance) performs substantially better than my prod environment (F2). Way faster cold boots and image loading. The dev instance was able to quickly render a 2MB hero jpg that consistently caused a memory fault on the prod instance. I've played around with different instance classes for prod (and eventually tinified the jpg), but nothing seems to perform as well as my dev standard environments F1. I have double and triple checked the instance classes in the GCP console for each project. At this point, I do not want to go to flex. I need to understand what is driving the inconsistency between the projects to be able to accurately gauge the appropriate GAE deployment for production.