I mean. If you use exactly the right gui framework and exactly the right coding patterns and you also implemented keyboard/gamepad support and not only mouse and you get in touch with the people who can tell you about tolk, and you do all of this early in your project and decide for accessibility at the beginning, it's not hard. But if any one of those things goes wrong you're screwed and it takes a lot of effort to undo mistakes and get into a position where you can even do it. Either you have really talented devs, you have devs who are willing to put in effort as a labor of love, or you have modders like Say the Spire where it takes literally months even though he basically had the game's source code and was building on top of a vibrant modding community.
But how this normally works out is-you don't care about beautiful abstractions. You don't write from the ground up but start from something that lets you write your game 10x faster, e.g. unity, where it's not accessible at all out of the box. You go write your one-off game event and you're not going "huh, how can I share this code", you're going "nothing else in my game is at all like this, I'll just hack it together". And for games, that's a perfectly fine coding strategy most of the time. But it does mean that unless you got every last thing I listed right from day one, plus the additional thing called "most of the GUI code is shared", it can literally mean going through the game and adding code to each event or whatever to make it talk.
And that's before we consider things that might not be adaptable without rewriting them to have an alternate mode, for example color-based puzzles that might now need to be entirely rethought for sound.
I'm a professional programmer. I think the fastest turnaround I've seen in my career on a feature that isn't like "add one button to a dialog" is a couple weeks across multiple people on the team. Programmers make $50/hour. A single programmer for a week needs to be paid $2000 to compete at the market rate. Accessibility will take several weeks and involve multiple people. My $20000 figure is considered cheap--it's what a small business pays for their small business e-commerce site for example. A game studio like this is probably running at below market rate, so their costs are lower, but it also shaves off what they can afford to lose in overhead--that is, if you're making $50 and your monthly living budget is $20 then working below market rate at $40 to help the blind people isn't a bad sell, but unless this game is staggeringly popular they're probably only doing $25/hour at best and at least in the U.S. most places that's really close to the wire as it is, so "here's something that doesn't make you money and also you can never stop doing it because it'll just stay with you forever every time you do a new thing" is borderline crippling for the company.
It's easy to say "this must be easy" but this kind of echoes a conversation me and a lot of experienced devs on here have with newbie audiogame devs, where it sounds like your small game is only going to take a couple days because it's small but you're still at it 2 weeks later. If you don't have significant programming experience, then whatever you're estimating as the timeline needs to be multiplied by at least 10 times, roughly. If you're going "well accessibility should only take a week", think more like 8-10, which is probably around what it would be here.
I want to be clear that I want to see more videogame accessibility overall, but at the same time what you all have to understand is this. When you go to a company and you say "make your thing accessible" they're going to say "and where do we get the money" and that's not actually wrong of them to do, since they have to feed people and want to continue existing as a company. Games are luxuries. As much as I don't see the point of The Last of Us 2, they probably spent anywhere from $100000 to $1000000 on doing it. If you tell your devs "we're making x blind accessible but this means we're paying you less" they'll just walk away, because while they might want to help the blind people they probably don't want to help us at the cost of their retirement, and it's not fair to ask them to do that for a luxury good that we want but don't need. If I was going to propose a solution to this, it would be to figure out how to get NFB/ACB/AFB/name-your-other-big-blindness-org to fund this work, probably starting with "hey unity, can you offer accessibility tools" or something, or just picking indie games like Slay the Spire and saying "here's $50000, have fun" or whatever else necessary. Someone like Sony can blink at sinking $100000 into a project like this for example, but most indie studios don't even make that in a year.