> Its 6m panels produce 1.5 gigawatt[-hour]s (GW[h]) of energy [per year]
or
> Its 6m panels produce 1.5 gigawatts (GW) of [power]
With some reasoning, I can land on the 2nd one just fine, but a little doubt lingers, because I've seen similar errors where I had to do the first substitution to make the numbers work.
A watt is just a joule per second, and a joule is a measure of energy, so I think that, in the same way, it's reasonable to talk about both "3100 joules of energy" and "200 watts of energy". I think it would probably be better to say "200 watts of power", but the other one isn't wrong. Is it?
Nearly every quantitative statement has an implicit denominator somewhere. Your server serves 50,000 requests per second, you say? Is that per core or per machine? Or maybe per rack? Occasionally there are implicit numerators as well: maybe those are 50,000 one-kilobyte requests.
Yes, it is. To use one of your examples, it's as wrong as saying you travelled a distance of 120 mph.
A journalist confusing units might have said something like "Its 6 million panels produce 3600 gigawatt hours," without bothering to tell you whether that's per day, per month, or per year. But it's perfectly correct to say that a panel or a solar plant produces some number of gigawatts.
† Yes, this depends on the type of panel and the location. Some panels are a little higher efficiency, some are a little lower efficiency, and the ones we use on satellites are a lot higher efficiency. And different places have more or less sun. Typical numbers are 1000 W/m² of "solar constant" nominal maximum insolation, 16%–21% efficiency (ratio of insolation to energy produced), and 15%–30% capacity factor (ratio of average power to peak power).