I don't pretend I'm an expert but I've been focusing on this sort of research during the past year and have found some strategies that have helped me. Many of these fall into the category of general good practices for all genealogical research but I think there is a case to be made for doubling down with "reverse genealogy" where you are often looking at an over-abundance of confusing and contradictory records.
First, keep good records. This should go without saying, but it's a point hard to overstate. In the case of reserve genealogy, you will be challenged with tracing individuals and families through name and location changes. Collecting details – even down apparent trivial facts – from each record or document you encounter will help you untangle identities as you move forward. Record even seemingly unimportant information – a street address, occupation, workplace, names of witnesses, officiants, even physical characteristics mentioned. And record your data in a spreadsheet or database format that will allow you to search and sort the information you collect. If you have a family with a lot of members and you really take a fine-grained approach to data collection, the list is going to get long and harder to manage. The ability to sort can be an especially powerful strategy, associating people with dates and locations. This is also a tried-and-true method for tackling brick walls and you should just start by assuming that you will hit a brick wall at some point: so why not build the spreadsheet right from the beginning?
Next, look at everything. You'll be surprised where something useful – sometimes revelatory – shows up. Sometimes it's one easily-overlooked item that turns out to be your family's Rosetta Stone. Obituaries, for all their faults, can be a great source for tracing extended family forward (i.e. lists of surviving members). Newspapers in general can be goldmine, esp. the turn-of-century society pages. City directories can provide insights, often in pieces which can be assembled to reveal a bigger picture. Common addresses can help identify family members, while occupations can distinguish those with common names from each other. A list of such details culled from year-by-year directory entries can help sift through other records you come across, separating your family from others, or finding your family even when they change names, locations or occupations.
Last, it's a hard rule of Jewish genealogy that spelling doesn't count. With reverse Jewish genealogy, names don't count. In some cases you need to work hard to link people with different names and in others distinguish among different people with the same name. Often the most reliable way of establishing who a person is, or isn’t, is through the correlating facts that trailed them along the way. But only by collecting and recording the related facts will you have the tools to trace and distinguish between those from your family.
Getting down off my soapbox… I hope this is helpful.
Lee David Jaffe
Surnames / Towns: Jaffe / Suchowola, Poland ; Stein (Sztejnsapir) / Bialystok and Rajgrod, Poland ; Joroff (Jaroff, Zarov) / Chernigov, Ukraine ; Schwartz (Schwarzman?, Schwarzstein?) / ? ; Rappoport / ? ; Braun / Wizajny, Suwalki, Poland, Ludwinowski / Wizajny, Suwalki, Poland