1) if you make it only applicable to smart phones, i just stop offering an uber smartphone app and now uber is website-only. if you apply it to "internet", as the original poster did, then:

2) companies like uber would be forced to shut down. you can say "cool, if they cant do it, their problem", which is fine, but a dozen of major issues pop up if something like 1/4 of the businesses currently propping up the stock market have to close doors or otherwise invest billions of dollars in phone centers or whatever they need.

it also raises questions about all sorts of businesses. another off the top of my head example: should 1password setup a call center where i can tell the operator what my new hackernews password is? is 1password exempt even if they have hundreds of employees and do millions per year? if yes, we have to come up with a bunch of murky criteria and definitions of what companies are exempt (across every industry, no less). which will, of course, cost a lot of time and money, just to surely be gamed. can we convince tax payers to foot that bill?

(this is also ignoring the approximately 0% chance that some sort of regulation of this sort gets pushed into law, against all of the extremely powerful tech lobbies. we dont even have ubiquitous right-to-repair!)

Then can do standard formulas like, will operations continue if the power is out, internet, smart phones, running water, phone lines, payment processing, etc, how long will service be down 1-3 days, weeks, months etc

If your store can't immediately switch to cash apply some modest tax increase. If people can't buy food for more than a week the extra tax is high. You might want to buy gas lamps and a "home" battery.

i am saying that you cant do a worldwide systematic review of everything that relies on the internet, and un-internet it.

if you have a realistic approach to doing so, i will eat my shoe.