There's a big difference between opinions that are merely taboo and opinions that get censored on social media. Not everything taboo gets censored and banned on social media.

Facebook/Twitter/popular social media as we know it barely existed in 2004, but I'm going to make a guess that discussing homosexuality wasn't banned on any of them (or whichever precursors of similar style and policies existed) around that timeframe. I feel like by picking that time frame, you're alluding to the speech standards that newspapers or TV imposed on themselves, and I'm very glad that social media doesn't impose that kind of standard on all of its users, but I don't think anything present suggests that social media is in danger of slipping to that standard. It looks like those dynamics are a world apart, and drawing these parallels is very misleading.

A bit of an aside, considering a taboo opinion which often prompts discussions about censorship online (an opinion that I think is bad and is taboo for good reasons, including that it's hard to detach from harmful calls to action, but I imagine it could be): talking about the possibility of differences in human capability in sex or race is taboo, but it's generally not outright censored in social media as long as it's not tied with calls to action like "#race war now". I feel like there's an overlap between those of the opinion social media is censoring too much and people outraged at James Damore being fired from Google, but I feel like people conflate him being fired from a tech company with the content moderation policies of social media. He was never banned from YouTube/Facebook/Twitter/Google services, but I get the impression that the volume of defense of him is linked to people thinking that did or will happen. It seems right to me that a large serve-everyone social media company would have different standards of its employees and the content hosted. (Silly example: I wouldn't expect a large social media site to ban a user for routinely posting that "$site is the worst", but I wouldn't find it surprising if an employee at that site was fired for that.) One could argue that Google went too far in firing him (and I'm glad those arguments are permitted in many places even though I disagree with them), but I often see it all bundled with the argument that this is a sign that social media itself is too censored and should be censored less, despite that it already permits him and that the enforcement line is still a world apart from him.

>It looks like those dynamics are a world apart

I guess my fundamental disagreement is that the two (censorship of taboos one deems undiscussable and censorship of taboos one deems discussable) don't seem so separate at all. Many of the same people want them, and for many of the same reasons. 1990s network TV was trying to appease the exact same public that Facebook would have tried to appease in 1990 if they existed back then and also had an attitude of appeasement. You, personally, have moderate views and sound relatively liberal in what you're willing to allow other people to discuss, but I'm not facing the decision of whether or not to appoint you as the censor: the issue at hand is whether or not we trust public outrage and the overall zeitgeist to determine what companies are afraid of being associated with, thereby determining what they censor. You can't set up a system with the idea that corporations will be reasonable or have a conscience.