The way I've always done it is to blend a sample of your finished product using weighted averages to get as close to your target as possible, while taking extremely detailed notes on details of your blending process. Then, send that sample to a lab that does .001% accuracy ABV testing. Note the variance from your weighted average blend, and then back add water/spirit/syrup to hit your target on the nose with the weighted average method. Then you could use the info on how much you had to add to the batch to adjust to target as an 'adjustment percentage' for future blending without sending to lab. If you want to be extra neurotic about it, you could then lab your adjusted batch to see if you were dead on. Making sure that your blending stocks (spirit, water, sugar syrups, etc...) are all the same temperature will increase accuracy with weighted average blending of complex liquids.
Unless you have a pretty legit lab setup, i'd be worried about incorrect data from non-laboratory-environment ebulliometry.
If you're small scale this method should get you as close as possible without spending dough on unnecessary testing. If you are making big batches of the stuff, it might be best to send every batch to lab and adjust as necessary if you want to be within acceptable % ABV variance.
Also, if you're worried about cloudiness it might be worth considering that citrus oils are pretty volatile and that by doing a 'demisting test' on your 'heads' coming out of the run you can pretty much exclude most of the oil that ends up clouding the batch. I do this on my gin and after about 1% of the run is collected the oils drop below the threshold that causes cloudiness at bottle proof. I just throw the oil-bearing 'heads' back into subsequent batches of macerate. I imagine that it would work about as well with your method for limoncello.