I'm completely blind. Regarding videos, less than a tenth of the videos that do have subtitles come with an audio description track for the blind, describing important imagery in the video that couldn't be figured out just by listening to the dialogue. While Youtube supports subtitles, sort of, it doesn't support audio description at all. The only online website that does is Netflix, and 4 movies on iTunes also have audio description. And...that's it.
Unfortunately, there is absolutely no way for a blind person to read any of the text on memes. As you can imagine, this leaves us out of about a quarter of Reddit. I don't even bother with Facebook these days, as about half of the things on my newsfeed are just images without description. The ideal web surfing experience would be if all important images and videos online had descriptions or labels.
However, this is harder than it at first seems. First of all, a person describing images needs to figure out if they're important, and why they're important. For example, if someone described all of the icons in a computer program with what they actually showed (thumbs up, paper clip, floppy disc, blue letter e, etc) a blind user would probably be quite confused. Instead, these icons get descriptions based on what meaning they're trying to convey (up vote, attach, save, internet explorer, etc). Also, most websites have lots and lots and lots of images. Assuming it takes at least 20 words to describe an image, describing every image on any major website would just drown a blind person in endless walls of text. Similarly, when an audio description of a video is being made, the describer can never speak over someone who is speaking in the video. So she has to carefully decide what the most important thing happening on screen is, and then squeeze descriptions of what's happening in around the dialogue in the movie. There just isn't time to describe everything!
So describing images and videos for the blind is, I think, much harder than captioning for the deaf. While I've never done it, I'm convinced I could provide adequate captions for a video; all I need to do is make sure they're in-sync with the audio, and write what people are saying. I don't need to provide any interpretation at all. On the other hand, describing for the blind is much more like translating, because you need to figure out why the image is important, and put it into words. As far as I know, there is no perfect, unbiased way to do that. That's not to say it's impossible; this
$1 may be the best example of describing images ever. But it did take pretty much a year of one woman's life to create.
Wow, that turned into a super long comment. I hope it gave you some insight into the different accessibility challenges for the blind, and why so few videos and images ever get described.