So, to have an IOT thermostat I have to give it around 350 ma @ 5 v (over 1.5 watts) 24 hours a day, 7 days a week? That's roughly 13 kWh over the space of a year.
It must be nice to design devices where someone else has to pay for the sloppy engineering.
Presumably the reason you would have an IOT thermostat instead of a regular (non-IOT) thermostat is that you want to be able to remote control it and have it collect performance data about the heating and cooling of your house. Being able to remotely turn something on when it is "off" requires it not to truly be off. 1.5 watts is not an unreasonable amount of power for this, especially if you include over time monitoring of temperature and that that the device needs to maintain a network connection. Where I live electricity costs about 11 cents per kilowatt hour, so a 1.5 watt load costs me $1.45 a year.
Good engineering is determining a plan that examines and balances the costs of various build options with a set of desired features our outcomes to arrive at an acceptable solution.
I suppose you could say that the engineering was sloppy and the load should only be about a watt bringing the annual cost down to about a dollar a year. What would the saving of that half watt cost in terms of design and manufacturing? Would the product now be too expensive to sell? Would customers even notice the reduced power consumption? Do customers care about a fifty cent operational cost annually? I would argue that extra effort to save the half watt is probably not worth it.
Setting aside the potential straw man, if the engineers who designed the thing considered the above questions then it was not sloppy engineering.