Furthermore, its not just that analytical grade reagents are 99.9% pure, and other reagents are (possibly) 99% pure. Its that vendors of analytical grade reagents certify that they are at the purity level they report. They have checked the reagent, against an aliquot of federally recognized standard, generally a USP standard, by a certified technician, performing an analytical method that the certifying body will accept. And they do check that the technicians at the company is performing to the correct standard. Such certification is provided to you when you purchase the analytical grade reagent, and you're required to reference the document in your own work, and produce the referenced on demand by a regulatory body.
Query: You've "performed some titrations/standardization(?) stuff and still got pretty accurate results." What does that mean? Are you testing foods for human or cattle consumption? Or are you testing raw materials, excipients or finished pharmaceuticals? Do you test fine chemicals, and guarantee your results to a certain standard -- example, you'd never report your results to greater significant figures than your starting material. You wouldn't report the result of your assay as 99.45%, if your reagent was only known to be 99.9%, correct?
If you have violated these rules, an independent auditor will want you to correct these shortcomings, before a federal auditor shuts down your company. And if you try to deceive people about your shortcomings, it isn't impossible for you to land in federal prison.