You've still got a cliff, just like in the means-tested welfare case; it's just that now it's on the consumption side.

Yes, that's what it's like (only the cliff is actually usually less steep under means-tested welfare). And you're also right about this:

In practice this would probably destroy the market for mid-priced goods - who wants to pay $1000/month just for an extra 10 square meters?

To clarify, I should say that my idea was that these subsidized or free goods and services would be so frugal that they would in effect not be an option to the majority of the population. Hence, it's not exactly the market for mid-priced goods, but the market for "low-priced but not extremely low-priced goods" that would get destroyed.

To your main point: since some people go down in standard, thanks to the fact that they by doing so they can get significantly cheaper goods, the average standard will go down. Now say that to get the average standard before this reform you had to pay 1000 dollars a month, but after the reform you just have to pay 900 dollars a month (because the average standard is now lower). Then those who want higher than the average standard will only have to pay more than 900 dollars rather than more than 1000.

The actual story might be more complicated than this - e.g., what some people really might be interested in is having a higher standard than the mean, or the the eight first deciles, or what-not. But generally it seems to me intuitive that if parts of the population lower their standards, then this should mean that those who want to consume consipiciously will also lower their standards.

Also, if you give out goods rather than money, you're going to have to provide a huge range of different goods/services, because otherwise there will be whole categories of products that people who legitimately can't work (elderly, disabled etc) won't have access to.

I don't see this as a comprehensive system: rather, you would just use it for some important goods and services: food, housing, education, health, public transport (in fact, the system is already used in the three latter; possibly housing too, though most subsidized housing is means-tested which it wouldn't be under this system). The system would be too complicated otherwise. Possibly it could be combined with a low UBI.