Hi!
First of all, congrats! I'm a high-school student and currently working on my own science fair project, so I can definitely empathize. I suppose I would try to make the conclusion correlate a little more with the hypothesize. For example, instead of "I hypothesize the colder temperature will result in a faster rate of growth, and therefore the crystal exposed to colder temperatures will form first." she could write "I hypothesize the colder temperature will result in a faster rate of growth, and therefore the crystal exposed to colder temperatures will not only be the first to form, but should also be larger than that grown in warmer temperatures. Over time, the differences in rate of growth should compound, causing the size difference to likewise increase exponentially." She could then go on to make a line graph showing the size of each crystal over time, with emphasis placed on the widening size-gap (assuming the data's available). Or she could, in her conclusion, emphasize how not only did the colder cyrstal form first, but it maintained its faster rate of growht over the entire time frame, resulting in a larger crystal at the end of experimentation.
On second thought, it may be a good idea to totally revise her hypothesis to place more emphasis on the cyrstals' sizes rather than which formed first, if that would be more conveniant for writing the conclusion. It's completely ethical so long as she doesn't change her actual prediction (i.e which temperature will result in the faster growing crystal)l.
As for the abstract, a good way to catch the judge's eye would be to emphasize the effects of her experiment on society. For example "Knowing the ideal temperature for borax cyrstal growth would decrease production costs. If the price of borax crystals were to decrease, the products in which they are used- insect repellent, flux, weed killer- should become more widely and inexpensively available as well."
Good luck at the regional fair