Well, I ran the numbers, and I was wrong. If I assume the average number of extra rapeseed produced per harvest is 26, then here is how it works out, first with the units...
rapeseed/sec * oil/rapseed = oil/sec = 26/810 * 40/2 = 52 canola oil / 81 sec
canola / sec * light oil/canola * solid fuel/light oil = solid fuel / sec
52 / 81 * 60 / 100 * 1 / 10 = 26 solid fuel / 675 seconds
canola / sec * pet gas / canola * solid fuel / pet gas = solid fuel / sec
52 / 81 * 10 / 100 * 1 / 20 = 13 solid fuel / 4050 seconds
26 / 675 + 13 / 4050 = total solid fuel / sec = 169 solid fuel / 4050 seconds
megajuoule / solid fuel * burner efficiency * solid fuel / sec = megajoule / sec = megawatts
25 / 1 * 1/2 * 169 / 4050 = 169 / 324 ≅ 0.5216 MW = 521.6 KW.
A greenhouse takes 400 KW.
So, you produce just 20% more rapeseeds than you need to power the greenhouse. After accounting for other overhead (the steam and electricity needed for the cracking, and for powering the assemblers and inserters to automate the process), you probably do not much better than breaking even for 10 greenhouses.
Once you move to greenhouse 2's you get 2.6 times the product for 2 times the power (and a bunch more space) so that's an excess of about 556 KW worth, or 69% more out than you put in. And, of course, there's the second level of fertilization with all of its attendant hassles which buys you even more production.
Still, on a space basis, they're about as efficient as solar panels, discounting the overhead. But they're a very large pain to set up compared to solar panels.
And you would have to have gigantic farms with maybe hundreds of greenhouses to produce enough rapeseed to produce all the plastic, lube, and sulphur you need in addition to being able to self-power the farms.