Blame the government (it's almost always safe to do so, and a favorite pastime for many people...)
EU MRLS tend to be written in ISO terms: mg kg^-1
U.S. tolerances are in the Code of Federal Regulations mostly as "parts per million" or ppm.
1 ppm = 1 mg kg^-1
Most of the time density is 1, so 1 kg = 1 L (note, this is a capital L not a lower case l)
Most of us work with g and mL, not kg and L. Also, most of my standards come in a purity something less than 100%, and a lot of them come in a salt form that has to be corrected for.
Some folks (biochemists most noteably) think it is nice to use molarity. It's always fun to convert a manuscript from nM steroid to ppb steroid
I've also been doing this awhile, and the first time I do a calculation, I write it all down, with the units, making sure they all cancel out nicely. Of course, working in a regulated environment, I would never ever do this on a scrap of paper.
We recognize that this is both a difficult undertaking and one where a mistake could affect results for weeks or months to come. Therefore, weighing, dissolving and calculating the stock concentration of a standard is the one activity that is ALWAYS witnessed and checked by a colleague, even for our non-GLP studies. Any subsequent mistakes in dilution would be caught in an audit, since all volumes used are always written down. Sometimes it is even possible to salvage something in this case.
Finally, fairly early on in method development/validation, a method SOP is prepared. These always have fairly detailed descriptions (weights, volumes of stocks and diluents used, exact calculations, etc.) of how to prepare stock solutions, fortification solutions, standard curve solutions, QCs, etc. One of my coworkers with a background in pesticide residue analysis actually takes this a step further and prepares worksheets to be filled in for each standard solution made. You spend a lot of time up front making sure all your calculations are right, but after that you just "follow the recipe."