The problem is in the real world, you can't perfectly optimize prioritization. In theory, you can run 5 4k streams on 100Mb. However, in practice, real time data can get messed up.

Before switching to 1Gbps fiber, I had 100Mbps and I frequently noticed that Zoom would lag when 2 people were streaming at the same time. There were enough situations where 3 clients simultaneously needed more data that introduced latency.

That's poor management by your network hardware. (could be your modem, router, ISP, etc) I ran two zoom meetings on my 1.5MBit DSL connection before I got Starlink and it worked fine. Or I could run Netflix and a Zoom meeting. Never did try to push it much past that. But on 100mbps, there's something else wrong there that you can only get one more connection than my shitty DSL line lol.

100Mbps fiber, or 100Mbps non-fiber? Because that sounds like an upload limitation caused by a limit much smaller than 100 or even 20.

It's the incoming videos that stutter. Not the out going video.

It’s not that but total bandwidth capacity is limited by constellation size, waiting for laser links between satellites, etc.

The bidding process had ISPs to declare what speeds they could deliver down to a minimum of 25/3 Mbps. SpaceX didn't need to bid at such a high performance tier if they didn't want to.

I agree that is overkill today, but if we are spending a bunch of money building out new infrastructure it makes sense to future proof it. The ISPs have ten years to complete their rollout, and I'd like to see it provide adequate service for at least a decade after that. Will 100Mb be overkill 20 years from now? Think about trying to use the internet of today on connection speeds that were common 20 years ago.

I'm a technologist in the media sector, I worked for a big ISP in one of the innovation teams and we were struggling to think of future applications that would actually occupy 1Gbps for consumers.

You might say "low compression, high quality video", true, uncompressed 1080p50 can be 3Gbps, but there's no consumer device available or planned that can handle raw video delivered in this way. You can easily compress HD video to 45Mbps and it's indistinguishable from the source to the average consumer. 4K at 150Mbps is easily done and again, most TVs and STBs cannot handle that kind of input. You'd literally need new silicon to be designed for most consumer devices to be able to do this and no one is starting that process.

There is an argument that low packet latency is important, but in reality more than 95% of people access their broadband over WiFi which negates most packet timing arguments.

There's arguments about VR, but all evidence is that VR is still not getting market traction, despite what Meta wants.

Sure, I can't predict 20 years from now, but there's nothing on the horizon (for consumers) that will stretch the current ultra-fast broadband for the average consumer. Power users are an exception and should not be the basis of business decisions unless they are paying the premium. We're fortunate that gigabit exists because of the competitive nature of the environment, it's not really about meeting consumer needs.

Pessimistically, if the next 20 years are anything like the last 20, I would expect most bandwidth growth to come from sloppy development and advertising, plus increased multi-device use, such that even if no single use needs that bandwidth lots of smaller amounts will add up quickly. Several hundred MB to load a normal webpage, before it starts streaming 5Mbps video ad, dozens of IoT devices phoning home orders of magnitude more often than the need to, on top of desired use of 50Mbps 4K TV streaming, and multiple 5Mbps TikTok/YouTube-like streams could easily make 100Mbps seem slow.

25mbit 4K ewwww this is why we don't yet have same bitrate of 4K blurays something like 128Mbps

Costs 5x more to deliver, few people would be able to consume it and 95% of consumers wouldn't notice the difference. You might be in the 5%, but most people you know aren't, or at least don't care enough.

Multiple studies by companies whose business it is to care, have shown that consumer engagement in viewing caps out at 1080p60+HDR, beyond that most people don't care. Sure, marketing creates hype and demand around bigger numbers, but when people are actually watching, they don't actually get more involved in the content above 1080p+HDR.