When I turned 32, I decided to learn guitar. I was absolutely horribly. Atrocious. Ear-piercing.
Four years later, I recently sat down in a jam session with a buddy who asked me how long I'd been playing for, and was shocked to hear that it had only been as long as it had. I wouldn't consider myself especially good at guitar, or any faster at learning than the next guy, but I've stuck with it, and made sure to learn something every couple of days, with the idea that if I learned a couple of things per week, then over time, those weeks would add up, and I'd know something.
I'm still a million miles away from being a rock star, but I could, realistically, play guitar in a band on stage, if I needed to. Learning songs has gone from taking months to a few minutes, on average (highly depending on the song). All that, and I didn't start til I was 32.
As a 36 year old web developer by trade, the skills are comparable. Learning compounds over time, and you'll be able to put pieces together more efficiently the more you do it.
The most relevant advice I can give is to just start learning, and the best way to start learning is to start doing. There have been plenty of folks who went from nobody to hirable by building a pet project and learning the language in order to do so. To work at a startup, or small business, "being able to code well" is often less important than "being able to code a project to completion." Build a blog, then use it to blog about your coding progress. Share your trials and tribulations, and post all the code you write on Github. Do that, and you'll likely be coding pretty well within a year's time.
To sum up, yes, you still have plenty of time to crack into the field. No, I wouldn't hire you (at the moment).
As for your niche market, my guess would be that your architectural background should have bearing on it. A buddy of mine is a restaurant owner, and also a developer. Unsurprisingly, he's building analytical tools for restaurant owners. He could likely have built the app for any niche, but his domain expertise means that he understands the market far better than even a much better developer.
At some point in your prior career, I have to assume that you dealt with some pain point related to poor tools in the job. Make better tools. Make them be more usable, more UX oriented, and more attractive, then sell them to other architects who were likely to have experienced those same pain points.
As for skills to improve, I would start with Javascript. For 'catching up', you can use the same language in both the front and back ends with JS, and you already know a smattering, which will hopefully ease the learning curve somewhat.
Either way, good luck. What I can tell you, as an average learner, persistence is really the only thing that you need. Just like exercise, or learning guitar, it all seems hard at first, and all gets easier with time. All that is required is to not give up early on, and you will most undoubtedly become better and better over time.