An old retiree, I've been extracting flavor from tobacco for several years. A NET fan, tobacco extraction became an enjoyable and rewarding hobby. To date I've performed hundreds of extractions on about two hundred different tobaccos/blends using numerous solvents and various extraction methods all in an effort to find/obtain the best flavors available that suit my personal taste preference. I've learned/discovered a few thing along the way but only recently did I discover one of the main reasons NET is so hard on wicks and coils (gunky).
It's widely known among NET fans that juice made from tobacco extract is extremely hard on wicks and coils, re-wicking and dry-burning after a single tank is common especially for those picky about flavor (myself included). What I've come to realize is that mechanical filtering, even at sub-micron levels, isn't the answer as filtering alone doesn't remove the cause of the problem. Many people have suspected the sugars contained in the tobacco and or casings/toppings were the culprit and while I agree, to a point, what I discovered was the waxes, oils, resin and other undesirables the extraction solvent leeches/pulls from the tobacco contributes to the problem -greatly-. For lack of a better word I call the waxes, oils, resin and other undesirable elements we extract "Lipids". Removing or reducing lipids is an important step and yields a much cleaner better performing, better tasting extract. I currently go through three to four tanks (9 - 12ml), before re-wicking an even then my coils look no worse than if using synthetic flavorings. I've only achieved this level of performance with extracts that used ethanol as the solvent and had the lipids removed or reduced via the freezing process. I currently filter at 1 micron but have played with sub-micron filtration although I'm wasn't overly impressed with the results.
PG, ethanol and VG will not freeze (become a solid), in home freezers but PG and VG will suffer a positive viscosity change (they become thicker), ethanol doesn't.
Hot ethanol is an aggressive solvent that extracts extra flavor but it also pulls considerably more lipids from tobacco than does PG or VG based extractions. If not removed these lipids can cause the extract to taste "off" and performance (wick/coil longevity), to be very poor. Exposing an ethanol based extract to freezing temperatures for a day or two causes the lipids to precipitate out while gravity/time causes them to coalesce/stratify in the bottom portion of the extract/container. Filtering them out is only possible while the extract is still freezing cold and fortunately ethanol is as easy to filter cold as it is at room temperatures. I even place the filter, funnel and catch container in the freezer prior to use so that the apparatus is at the same temperature as the extract before filtering takes place. I haven't tried to remove lipids from a PG based extract yet so don't know how effective the freeze/filtering process would be in doing so but considering the potential benefits I'd certainly attempt it. Since it might take some time for the cold (thicker), PG to filter through I'd leave it in the freezer till it was finished doing so.
EDIT: For the freeze filtering process I use filter felt or a coffee filter to separate/filter the gelled lipids. Filter felt is ideally suited for this purpose due to its fuzzy texture and thickness. Final filtering (In my case through 1 micron borosilicate glass filter paper), is performed after the extract has warmed to room temperature.