I found his units to be significantly more clarifying than yours. There's nothing wrong with having units of time (or distance) in both the numerator and the denominator, depending on the context. For instance, the Hubble constant is commonly expressed in km/s/Mpc, and astronomers purposefully don't cancel the units of distance.

In this case, there are two distinct timescales involved: the scale on which the energy is used (an hour or day) and on which it's generated (a second). It's ok to keep track of those separately.

There's nothing "wrong" with the units, besides that it is easy to get confused, evidenced by GP calling the turbine "7MW(h?)".

If you want to compare, however, you will have to convert to common units, and I at least find it easier to use watts and joules than watt-hours/second and watt-hours. By all means convert to stationary-bike-hours and big macs or whatever unit at the end but calculations are easier in the SI units.

Hubble's constant is expressed like that because it is a nice unit for the math.

> In this case, there are two distinct timescales involved

Three really: seconds, hours and days.

8 kWh per day = 8 Joules per second ⋅ hours per day ~= 330 Watts

Sometimes you even get four time units in one go. e.g. "I averaged 8 kWh per day last year"

It's madness.