Not saying you're wrong, but where does it state this? I didn't want to download the full report to check myself as it asks for email/name etc and so it'll have to wait until I get home. But all I could find on that was:
"Since 2010, the National Solar Jobs Census has defined solar workers as those who spend at least 50 percent of their time on solar-related work."
Which could be taken either way.
As an extension, are you saying they define "solar related work" as working > 50% of your time for any company which is any part of the solar industry? And if so, could you tell me where they state that?
So it's fairly detailed.
Having said that I think it would be fair to count home installations anyway. If we are talking about the job creating benefits of power generation methods then we should talk about jobs per installed Megawatt both in terms of initial roll out and ongoing maintenance.
There is no option to use coal generators at home, although I believe coal heat was not uncommon in the last century. If such a thing was common today then "residential coal boiler worker" could also be counted as part of the coal industry. Coal is not a technology that lends itself well to home installation and so there are fewer people involved in it.
One of the benefits of solar is that it can be installed on individual homes and this will necessarily employ more people to support all the individual installations.
It's equivalent to counting every housebuilder in the country as "employment in the automotive industry" because they install garage doors on some of their houses.
Solar's economic advantage is that it doesn't require many jobs. The panel on your roof just sits there producing electricity, with no need for regular shipments of coal to be transported to it by train... It shifts power production from an ongoing cost to a fairly small capital cost.