Does the camera face the x axis when the yaw is 0?

First, let's bring a little bit more context into your question so that we know what your are actually talking about.

We can assume that when you say

so I’ve been reading about the camera in learnopengl

that by this you are specifically referring to the chapter called "Camera" in the https://learnopengl.com/Getting-started/Camera tutorial.

Under the sub-section "Euler angles", there is the image which you are also including in your question.

If you read a little bit further, you'd see the following definition of the direction vector, which that tutorial later uses to build a lookat matrix:

direction.x = cos(glm::radians(yaw)) * cos(glm::radians(pitch));
direction.y = sin(glm::radians(pitch));
direction.z = sin(glm::radians(yaw)) * cos(glm::radians(pitch));

So, just by looking at that and doing the math, we see that direction will be (1, 0, 0) when yaw and pitch are 0.

But, if we read further, we see this paragraph:

We've set up the scene world so everything's positioned in the direction of the negative z-axis. However, if we look at the x and z yaw triangle we see that a θ of 0 results in the camera's direction vector to point towards the positive x-axis. To make sure the camera points towards the negative z-axis by default we can give the yaw a default value of a 90 degree clockwise rotation. Positive degrees rotate counter-clockwise so we set the default yaw value to:

yaw = -90.0f;

So that is the answer to your question: In the context of that tutorial, by simply its own definition of the direction vector, it will point to (1, 0, 0) when both angles are 0. And to counteract this, that tutorial will assume that the initial value of yaw is -90.

Sure, they could've just used a different formula for the direction vector components to yield (0, 0, -1) as the result when the angles are 0, but they didn't.

I think you might be confusing definitions of camera space and projection space. -z-forward is how the projection matrix is defined, whereas the above is how the view matrix is defined, when you multiply them, you get the effects of both (a perspective corrected final image that creates the illusion of depth, as well as orthogonal world space coordinates that you can properly translate and rotate in).

If you want to define it to be -z forward:

We replace the trig for yaw, and pitch isn't affected:

direction.x = sin(glm::radians(yaw)) * cos(glm::radians(pitch));
direction.y = sin(glm::radians(pitch));
direction.z = -cos(glm::radians(yaw)) * cos(glm::radians(pitch));

You can now use Yaw = 0.0f and the default camera will be looking down -z.