Another field is quantum sensing, from highly scientific (detecting particles/matter/etc.), to highly down-to-earth (e.g. military). Some properties of QCs can be used to improve the precision of quantum clocks.
There are so many exciting fields and applications, where progress is made and will be made before general purpose QC is developed on which you'd run Shor's and Grover's (if ever, in our lifetime). The truth is that QCs are best at niche unique applications related to quantum mechanics, and not so good at general-purpose computing.
Sure, pop media is not the source, it's being fed by industry spokespeople, who are ultimately concerned about funding. Many people in the industry do believe that quantum winter is coming. But it's kind of a self-destructive cycle: to continue progress we need funding, to get funding we make promises, to make them efficiently the industry gravitates towards easy-to-blow-your-mind topics like breaking encryption or solving traveling salesman, but chances are we won't have a QC capable of practically useful Shor in the decade or two at best, and by that time funding may decrease significantly because, well, "we can't Shor yet".
Examples like PsiQuantum raising almost 1B from partially public sources, and then delaying their ambitious plans does not help anybody. If they fail, the general public and the policymakers would, again, confirm their suspicions that "QCs are a scam".