This is an example of a tenant cost, not a landlord cost. In fact, the landlord may not be able to increase rent as much in the future due to the utility cost increase, because the tenant has less disposable income.
> If the tenant makes $1M/year it won’t really affect what the landlord charges them.
I don't think an extreme end of the scale is particularly useful to consider in this case; most tenants don't make $1m a year, and most people who make $1m a year don't rent.
This is relatively easy to resolve, anyway. If "Costs are frequently passed on directly" as you say, then you should be able to make some predictions about landlord costs and rent and then see them replicated in the market.
An example would be:
Many landlords have mortgages. An increase in mortgage rates in countries with non-30-year mortgage rates is an increase in landlord costs. We can expect a doubling of mortgage rates to have a significant impact on landlord costs, and thus a significant impact on rents. Do we find in countries with non-30-year mortgage rate rises that rents go up significantly more than usual with significant mortgage rate rises, or not?
You can also think of an example more to prove whether rents go up when tenant incomes go up or not, such as:
Do rent increases have a correlation with tenant income increases, or are they unrelated?
In my country (New Zealand), there is ample evidence to support landlord costs not generally being passed onto the tenants (I say generally because there will be specific instances where this isn't true).