Well, CO2 in the atmosphere (and amount of carbon above ground in general, a lot has been brought up from below in the last 100 years) and plastic pollution come to mind. Ocean pollution. Bio-diversity. Less or no weapons of mass destructions. No climate change yet, e.g. in the western US water from rivers was less of an issue and could be distributed more freely.

There were a lot of extremely dirty industrial processes which we are doing much better now (in the developed countries at least), superfund sites in the US and their equivalent in other industrialized nations as the known locations, but it did not reach the global scale yet, plus, what they buried or just left in the ground where the factories spilled it is still there now for the most part.

I think that in quite a few ways we've gotten better, sure, but partly by kicking the can down the road instead of actually solving issues such as sustainable energy or water use instead of drawing from and heavily relying on exhaustible stores, as well as not having closed loops with the things we produce.

Hugely more pollution (increased exponentially from 100 years ago as shown in graphs).

Higher temperatures (climate change, not the same as pollution).

Worsening infrastructure (the US, for example, is turning into a developing world country in many areas) as opposed to the era of great infrastructure projects, from scryscrappers and railways to the interstate system .

Lower/flattening productivity increases (as opposed to continuous productivity booms 100 years ago).

Lower/flattening invention rate (the low hanging fruits ended circa 1950-60).

More surveillance (in fact more, even at solely the state level, than even Stazi had, and with infinite potential for targeted surveillance).

Thining middle class (as opposed to incresing middle class 100 years ago).

Hugely increased inequality.

Crumbling pension systems (as opposed to pension systems being developed 100 years ago).

There are also a few tricks such "we're getting better" treatises use:

1) use the progress in countries starting from a very low bar to offset regressions in more developed countries (and since the third world countries have higher population, use an average state to shove it to the face of doubters who see their own lives and those of their countries/societies get worse and worse). This is kind of like averging Jeff Bezos' and mine wealth (kinda: what it does it says since e.g. 1 billion people went from $1/day to $2/day this is a huge 200% increase, who cares is another billion went from $100/day to $80/day and falling...).

2) take official measurements (made to make states and organizations affecting change look good) as the ground truth

3) when convenient, compare with say a century back, not 30 and 40 years back (e.g. when people didn't even have the notion of a pension, not when you expected and got one).

4) ignore any qualitative metric.

Not a direct answer to your question (also 1921 isn't exactly a local optimum considering it's just after the end of WW1 and on the precipice of the Great Depression) but if you take away a "primitive" tribal community's land, charge them for living in it and pay them a starvation wage to work the land for you, you've increased their economical wellbeing by almost every metric.

Additionally the metrics being called out are things like the World Bank's poverty statistics, which applies fairly arbitrary definitions of income-based poverty that in many cases would identify people otherwise labelled "food insecure" as "not in poverty". The definitions are also applied globally and by their nature adjusted over time, which can skew the results. And even if all of these limitations are ignored most of the "improvements" are localized in China.

I'm not going to defend primitivism because I think the problem today is mostly a matter of distribution and resource inefficiency but for a metric that has gotten worse over time (though again not in the 100 year view for obvious reasons) consider leisure time. Hunter-gatherer societies spent considerably less time on sustaining themselves than the 40+ hours per week the average modern human spends working a job, doing household chores and shopping.

The problem is that a lot of the ills don't neatly fit into metrics and even when you try to do so, you have to rely on self-reporting for a lot of them. For example for most workers industrialization massively increased alienation from their work, remote work (but also on the opposite end the need to relocate for career advancement) has drastically increased social isolation, enclosure and the increasing commodification of leisure activity also have take their toll, and social media has increased the spread of disinformation, replaced social interaction with public performance and has been a massive cause of social anxieties.

And of course the thing is that a lot of the problems we have can't just be fixed by a clever improvement because their causes are intrinsic to the system. You can try to reduce overproduction and improve recycling but the best consumer action to help the environment is not to consume and the best industrial action to help the environment is not to produce. But considering how much of our economy hinges on demands people didn't know they had for things they didn't know they need (not just production but also sales, advertising, financing, sales and advertising for the financing, debt collection, sales and advertising for the debt collection to the companies doing the financing, insurance for the companies doing the financing, sales and advertising for the companies doing the insurance, and so on) that just translates into "killing jobs" and "ruining the economy".