We got to this point because the early small rockets are trivially easy to raise, and when the space race started in earnest we just threw money at the problem. The Russians didn't build and transport their Soyuz vertically because they couldn't afford to, whereas we could afford to do almost anything. SpaceX took a page from the Russians and became the only US launch provider that doesn't build their rockets vertically, just standing them up prior to launch, and not coincidentally is by far the lowest cost provider.
If they were extremely confident in their engine start, they could lift the nose up with just 30 SuperDracos, and they're already going to use eight of those in the Dragon capsule just as abort engines. They don't need to do that because their rocket is small enough to raise vertically right at the launch pad, but as rockets get bigger that operation becomes much more difficult and expensive, and eventually impractical.
So, a question: How many Orion capsule abort motors would it take to raise the entire SLS to vertical while its fully fueled? About six. If you use twelve you could lift the entire rocket, including the SRBs, and flip it vertically in the air. The abort motor is the tiny little thing on top of the stack. Lifting and fipping are infinitely easier than getting such a vehicle to go supersonic, and then getting it all the way to outer space.
Wrong on all accounts. When is comes to impractical, it is your idea.
a. OSC and Delta IV are horizontally integrated vehicles.
b. Wrong on the Dracos. To lift the whole vehicle requires nearly the same thrust as the main engines. For example, using a Falcon 9, the thrust of all nine engines would need to be directed down. And to maintain control, some of the Merlin engines would need to mounted near the nose. Not puny Dracos.
c. Lifting and flipping is not easier than vertical assembly building.