Why must you laugh at my back end?

That's one, and it's a big one.

The other, even bigger benefit is that other people will do things for you. They'll write plugins for you. They'll develop new versions of the framework for you. They'll do crazy experiments in corners of the framework design space ("Hey, can we make Drupal work with a CouchDB back end through a series of proxies?") and report the results in obscure Stack Overflow posts that will get indexed by Google. They'll find and fix bugs and security holes for you. When you ask one of them about some problem in your system, they'll have a fighting chance to give you useful advice -- even if they've never actually seen your system -- because, hey, all Rails apps are kind of alike, and all Drupal sites are kind of alike, and all node.js installations are kind of alike.

The fact that this is the major benefit of using a common framework is also the reason why the poor OP feels so lonely. The whole point of a common framework is to have lots of meetings and rallies and beer bashes and hackathons and IRC channels so that everyone can swap ideas and recruit each other and basically work in a big team. They generate their own buzz, because the buzz is part of the value. But hand-rolled personal systems don't get that kind of buzz, because -- to paraphrase the ever-paraphrasable Tolstoy -- all standard web frameworks are alike, but all nonstandard web frameworks are nonstandard in their own way.

Obviously not every software tool discussion is as exaggerated as that, but even if you're arguing Django vs Rails, there really are significant differences that could affect your work in dramatic ways (e.g. one framework has a plugin that would take you a week to write, while the other doesn't).

look at wordpress, joomla, drupal

I have a theory that people act this way because of evolution.

--------------------

Imagine you live in a tribe of 10 people, and there are 5 different ways the entire tribe can use to hunt animals for food. All 10 people must hunt together, and each of the 5 ways are equally successful.

After an animal is killed, the tribe determines who was most responsible for the kill, and he/she gets to pick and eat the tastiest meat before anyone else. Others' still get to eat, but it doesn't taste as good.

If all 10 hunters are equally skilled, each will get tasty meat 10% of the time regardless of the hunting method. But humans are not all equally skilled. Some can throw spears better than others. Some can run fast. Some can sneak up quietly. Some can climb trees. Some can wrestle a wild bore.

So what happens? Before a hunt, if you're a fast runner, you say "Let's chase after an animal!" If you can climb trees, you say "No, let's drive an animal up a tree, and climb up to kill it!" You push for the option which gives you the best chance eating tasty meat, and do everything you can to stop any other hunting methods.

Things get political. Someone says; "Running after animals is too much work! Lets climb trees instead!" You say; "Bullshit! Climbing up trees is for pussies. Real hunters run after their food! Plus, we have to see your nasty balls every time you climb up!" You support your cause, and do everything you can to demonize and discredit other options.

Through politics, arguments and fights, you might be able to sway the majority opinion of your tribe in your favor. "Yeah, I guess I don't really like climbing trees either; I'm sick of seeing that guys balls too. Let's run instead!" Now, your entire tribe runs for food, and being the fastest runner, you get tasty meat more than 10% of the time.

--------------------

So my theory goes, when you see someone arguing about something pointless like Windows is better than Linux, even though both are pretty good, what they're really trying to do is sway the opinion of the entire world in their direction. If every company dropped Linux in support of Windows, there would be a lot of Linux sys-admins and programmers out of work. Sure, they could learn Windows, just like a slow runner could learn how to run fast, but they would be at a disadvantage compared to seasoned Windows admins.

The same thing happens with programming languages and APIs. These can be very volatile, going from extremely popular and hot, to nobody wants to touch them in a few years.

When you see someone arguing about what color is better, they're using the same algorithm, only it's firing on something seemingly trivial. But is is? Perhaps if you convince enough people that red is better than blue, the next Macbook Pro you want to buy will end up red instead of blue. Unlikely, but possible...