That's not a bad thing for a sufficiently well designed test. Here are examples of decent tests which I taught to. "Teaching to the test" and "teaching calculus" were basically identical.
"Teaching to the test" and "teaching calculus" were basically identical.
I'd argue that if that's how math is being taught, there's a problem. "Teaching to the test" taught me to hate math in high school. In middle school, we were given a brilliant teacher who was allowed to teach however he wanted, and we left 8th grade understanding trigonometry better than most of the high schoolers we were with.
Have you read Paul Lockhart's "A Mathematician's Lament"? That highlights the problems fundamental with the current math curriculum. From the tests that you linked me to, with all due respect, it doesn't look like you're doing anything different: you're spitting out formulas to be memorized. That's not math: that's mechanics. I hate that students are taught like that.
The mechanics are important, and calc 1 is the course in which they are taught. If the mechanics of algebra and calculus are not second nature to you, everything that follows will be almost impossible. Similarly, when learning programming, you need to learn to write syntactically correct code, and when learning a foreign language you need to memorize vocabulary. It isn't fun, but it is necessary.
Later courses are more focused on mathematical reasoning, and tests reflect this as well. For example:
These exams are also pretty tough to game (even though they are standardized); the easiest way to teach to the qual is to teach how to reason about real/complex analysis and algebra.
You don't seem to agree with the curriculum. Lockhart's article claims that courses like Calc are not well suited for elementary school math (and I agree). That's an orthogonal issue.
The calc tests do a decent job of measuring what they are trying to measure: how well students understand the mechanics and concepts of differentiation and integration.