You estimate their "unreasonable" driving at 40 mph, but then your travel plan assumes driving at 60 mph.

Charging adds around 7 hours to a Seatle to Miami (3,300 miles, 48 hours driving, 7 hours charging) or San Diego to Rhode Island (3000 miles, 44 hours driving, 7.5 hours charging) in a Tesla Model 3, or about 10 and 12 hours charging for the same trips in an MG4.

That's not nothing, but it's also not adding an unreasonable amount for someone doing that trip in 4 days. It's a forced 30-40 minute break every 2-3 hours, and turns a each 12 hour driving day into a 14 hour driving + charging day.

CLR is an acid-based rust remover

It's reasonable to consider capability, but if you're going to do that, you also have to factor in the reduction in demand for capability due to chargers in homes, which means that ~700,000 cars no longer need to be charged at public facilities most weeks.

I think it comes down to how much is "don't drive enough"?

As someone with an EV who lives in a small regional town, 100km from the nearest city, 100km in the other direction to work, who drives 250-350km (150-200mi) per day on normal days, with 600+km (370+mi) days every month or two when visiting folk, I get by with overnight home charging each night, with the need for fast chargers limited to the longer trips. We've owned this car for over 3 years, and it currently has 133,000km on the clock, so we use our car significantly more than most people.

Every morning I leave the house with a full battery, and return home with between 40% and 5% battery each night depending on how far we went that day, and what the weather is like. It takes a maximum of 6-7 hours to charge overnight, if the battery is at its lowest, so every morning is a "full tank". The cost of charging, even paying the premium for "green" energy sources, and without having solar, is still ~1/3rd of the petrol cost we were paying previously.

In three years of having this car, we've never had to cut a trip short or make any serious deviations due to battery levels, except on the first road trip after owning it for less than 2 weeks and still learning how to drive electric. On the longer trips (500km-1000km), we have been able to synchronise our charging stops as food stops, toilet stops, or quick leg stretches, typically under 20 minutes. On the rare occasion, only slow chargers were available in the town we stopped in, but not often enough to be remarkable. Granted, we haven't yet done the 2000km+ road trips that we were planning (for unrelated reasons), so we don't have real-world experience of that, but when planning that, it wasn't going to inconvenience us in any way.

I agree that if your daily usage cycle is more than 300km, and you're at home for less than 7 hours a day, then an EV is currently not for you. Anyone who has a usage cycle like that will have to wait for future options which might better suit them, and will have to remain using an ICE car for now. Also, if you're regularly (more often than once in 3 months) doing time-sensitive trips in excess of 1000km, an EV may not be suitable depending on your local charging infrastructure.

However, anyone with a dedicated parking spot which can have a charger installed, who drives less than 300km per day most days, and is home for over 7 hours per day most days, should be able to use the current technology to improve their lifestyle rather than inconvenience it.

When we bought the EV, we held onto our ICE car as we didn't know how well it would work out. In practice, we don't use the ICE car any more, and we're selling it.

Framework is probably a decent counterexample - laptops are similar to printers in most ways that you describe, apart from consumables, and Framework seems to manage building and selling a range of field-repairable, open-standards hardware for laptops, including the spare parts and upgrades..

Surely Czechia, Slovakia, Slovenia, and Croatia would all count as fairly recently having broken away from the country they were part of?

Or "that thing we promised turned out to be harder than we said, so we're going to just not deliver it, and claim it's because the next version will be even better"...

Likely, the real answer is either "we made a deal with Mazda which says that only Mazda is allowed to sell this feature to you", or "we're in a commercial/legal battle with Mazda, and withholding this feature from their customers is a negotiation tactic for us".

The relevant car analogy for their claim is: you bought a car for $10k, and used 4 days and $400 in gas to drive from LA to New York. You've spent $10,400, however, you've can now claim that assuming $50/day for a rental car, you could have spent $600 to get from LA to New York.

(The claimed training cost is based on multiplying the number of GPU hours used in the last training run by the approx cost to rent a GPU for an hour.)

They said it didn't need the latest and greatest Nvidia chips, because it ran on older, less powerful, cheaper Nvidia chips.

They're asking for the difference between a McDonalds (bring the food I want to me) and a buffet (I go around and find the food I want).

There are scenarios in which this can happen, but almost any company which is large enough to eventually cross that threshold is probably a net benefit for society if it was broken up into smaller companies...

My suggestion would be - average revenue over the last 5 years.

If the average salary of the type of tech who would be doing this is around $125k (from a quick google), then $300-$3000 is a rough translation of "between half a day and one week of person-time" per VM. And while spending a week on a single machine for the migration seems like a lot, if you have to do testing, shutting down interacting machines during an outage, pre-move backups for rollback, etc, I can see how it can easily take over half a day per machine...

A team I know recently took over a year to move a bunch of VMs from a datacentre which was being shut down to a new one, without changing the infrastructure in any significant way.