But I agree with you.
If anything, PHP has proven it works in this regard.
PHP works, but imho starting a new company or career with PHP is a mistake.
Why would it be a mistake?
On the one hand I have a few examples of huge, successful projects written in PHP and I know they are still around 20 years later and haven't been rewritten in another language. I use projects in PHP on a daily basis (Nextcloud, Rainloop, Wordpress) that work well and have been maintained for years. On the other hand I have a few comments calling out PHP without convincing arguments. Something is wrong.
I practice Python and Javascript daily, I've written PHP code 10 years ago, still use some of my code. I occasionally update PHP code on a website I manage. I've seen good things in the new versions.
Should I begin a new project today, I would not rule out PHP that fast. I like Python and Typescript, but PHP projects are often a breeze to install (unzip, done) and survive system upgrades and reboots easily as long as php-fpm runs. Projects in other languages need more babysitting. I know, I also use web software in other languages, including one I author.
PHP is dead simple to write too.
I would probably go for nodejs for something that requires Websockets or SSEs (or a compiled language). Otherwise, I'd seriously consider PHP because it does not require setting up something like uwsgi or celery, write reverse proxy rules, write init service files or running Docker. Put the files in www and it's done. Also no code for managing routes, the file system organization does this for me. No memory leaks, everything is thrown at the end of each request. The architecture is robust.
