If I'm making money or even pursuing a hobby, $300/yr is completely worth it.
But FWIW, I'd be happier about having $40/month-billed-monthly and $300/yr options.
Milking money out of me on a regular basis, adding to a line-item in my budget, feels predatory - as a developer, this turns me off because it doesn't actually benefit me in any way - I'm paying for your inability to fund yourself through multiple iterations of software development cycles. Put a fixed price on the product, charge for upgrades when - and most importantly: if - they happen, and it would be a done deal in my case. But right now, I don't want to have to deal with your regular charge against my credit card. Too much hassle, and you haven't been in business/released enough software, long enough for me to believe its going to be worth a year of charges.
What I'm trying to say is that your "return" from this product is obviously contingent on how much you use it/what you make with it. If you have 95% confidence that you'll produce enough to net a positive ROI over and above the fixed price, then yeah it's more convenient to buy it once. But especially for a new product like this, I much prefer a subscription model because I have more control over my costs and I can't yet anticipate my true benefit. It's a win-win: software maker gets to push frequent updates and ensure a constantly improving experience, and user gets to check-in frequently and re-consider if they're not getting enough value.
I haven't tried PencilCase yet, but it looks like a great product with solid implementation. Good luck to you guys.