I'd argue that with more than two people in a household and everyone switching to streaming services, you'll need to offer at least 400-500GB data caps for fixed wireless 5G.
Several of my renters are single, and their streaming usage for an individual is ~450GB/month.
You need 400-500GB to accommodate who? According to Comcast, median Xfinity data usage in 2H 2017 was 131 GB: https://www.xfinity.com/support/articles/data-usage-average-.... That's going to be squarely within the range of 5G (analysts are guessing around 200GB caps at first). Also, expect to see video content be zero-rated (it makes technical sense--watching Netflix from one of Netflix's co-located CDNs puts a lot less load on the network).
Remember, 5G doesn't need to be an alternative for everybody to wreck the ability to raise prices to compensate for lost TV revenue.
> According to Comcast, median Xfinity data usage in 2H 2017 was 131 GB
Given their recent push for data caps (and make money off of overages), I'm sceptical of their numbers. They have a pretty obvious incentive to misrepresent (if not understate outright) the value to make their caps appear generous.
Also, is median (same as 50th percentile, right?) really the right statistic to consider? I'm not sure any one number is sufficiently meaningful, as the shape of the graph can be particularly informative.
More importantly, how has that changed over time, as compared with the penetration of 4K TVs versus 1080p ones (or the availability of 4K content or some similar metric)?
Conversely, I think mobile data is also more likely to be significant as competition in multi-person households, even with the caps you mention, because those have a chance of being per person/handset (absent family/share plans), while Comcast is likely to be one cap per household no matter how many people/screens stream at one time.
Several of my renters are single, and their streaming usage for an individual is ~450GB/month.