I tested a UK-based hosting provider on a Magento 2 project this month and the numbers were hard to ignore.
Time to first byte dropped from around five seconds to roughly 800 milliseconds. No code changes. Nothing refactored. Purely a modern hosting stack underneath the same application.
Why This Is Worth Saying Out Loud
When a store is slow, the reflex is to blame the code. Somebody proposes a refactor, somebody else proposes a replatform, and six months disappear.
Sometimes the application is fine and the floor it is standing on is the problem. TTFB in particular is a server-side number. If your server takes five seconds to start answering, no amount of image optimisation on the front end will save you. I have made this point before about hosting bills and caching, and it keeps being true.
The Order I Would Check Things In
Measure TTFB before you touch anything else. If it is bad, your problem is infrastructure or caching, not JavaScript. Fix that first, then look at the front end, because only then are you measuring what you think you are measuring.
Skipping that step is how teams spend a quarter optimising CSS on a server that was never going to be fast enough. It is the same trap I described in what Core Web Vitals actually mean on a Magento 2 Luma build.
Honest Caveat
These are early results. I am still benchmarking under real traffic and load, and I will believe the numbers properly when they survive a busy week. But it is promising, and it is the sort of gain that costs a migration rather than a rebuild, which makes it worth testing.
If you are running Magento 2 and want to know what this setup looks like on your store, ask me and I will walk you through it.