Yes, you've complained about it before - more than once. (At least, someone has - I don't specifically remember if it was you.)
And it's a valid point. The units are wrong for them to be compared.
And yet, it's only half of a valid point, because the only thing people are doing with that ratio is comparing it to the same ratio for other years or other countries. The number can be dimensionally wrong and still be useful to use in that way.
Why would the number be useful to use in that way? Well, it seems intuitively reasonable to say that the bigger an economy is, the more debt it can carry without the debt burden being any more of a strain. The US can carry more debt than Vanuatu can, and not because the US government is more credit-worthy. So scaling the debt by GDP makes some sense.
And, in fact, dimensional analysis can make sense of this. Let's say the GDP is $10 Trillion. Well, it's not really $10 Trillion, it's $10 Trillion/year. If the debt is $30 Trillion, then the ratio is $32 Trillion / ($10 Trillion/year), which equals 3.2 years. That's how long it would take to pay off the debt using 100% of the GDP (which of course never happens, but it's still a measure of how much of a load the debt is). So when you see them report that debt is 320% of GDP, what they're really saying is that the debt load is 3.2 years.