Somebody asked me how I derive the numbers I use in these reports. Here is my explanation:
For example, if I search on Crozier with no first name, I get a list of all profiles with Crozier as the Last Name At Birth. The current total of Crozier profiles is at the top of the page, a little to the right of centre.
Next, I click on the unsourced link, and that gives me a list of Crozier profiles with the {{Unsourced}} template applied. I just count them up. (With more popular surnames, I have to page through the list, and hopefully not lose track of how many pages I've gone through before I hit the first entry with a different surname, and then multiply the pages by 200 and add the number on the partially full page to get the total.) This number won't be accurate, for reasons I've explained above, but hopefully it will get more accurate as time goes on, as people add the {{Unsourced}} template to more profiles.
Next, I go back to the Crozier genealogy page and click on the unconnected link. Then I click "No" beside "Limit to Open Privacy Level" to get the number of all unconnected Croziers. That number is accurate, at least as of Monday morning. (I think that particular script only runs once a week.)
To see the Croziers on ThePeerage.com, I go to the home page there, then click on the Surnames button, then click on "C", then I type Command-F (Control-F for Windows or Linux) and type Crozier into the search box in my browser. Then I click on the highlighted link to see the list. Darryl includes both Croziers at birth and those who married Croziers, so I follow each link, and enter the name, dates, and ID number into my spreadsheet, keeping the living people (including those born less than 120 years ago with no death date listed) and those for whom Crozier was not their Last Name At Birth separate, so I'm only tallying deceased Croziers at birth. I also have columns for sourcing level, whether a profile is connected to the main tree, whether there's a picture of the person on their profile, etc. I also use IF formulae to put a "1" in the appropriate sourcing level column, so I can get totals of how many profiles are at each level. Then I have the spreadsheet do all the number crunching and chart drawing, setting the charts to use WikiTree-ish colours just to look better, and try to get the reports posted soon after the end of the month (although fairly often, Real Life interferes with that).
For Wikipedia, some surnames have articles like Crozier (surname). Others have more elaborate titles like List of people with surname Miller. Others simply have something like West (name). Others are more simple still, like McNair. The least common surnames have no article at all, and I just have to search for anybody with that surname. Assuming there is a page for a given surname, I work through it the same as with lists on ThePeerage.com, with the difference that there are also fictional characters to separate out.
For a couple of the measures, I use Aleš's reports:
https://wikitree.sdms.si/default.htm?report=srch1&Query=LastNameAtBirth%3DCrozier+unlinked&MaxProfiles=500&Format=
gets me a list of Crozier profiles which aren't linked to any other profiles. I have to say that this measure has come down a lot in recent years. A while back, most of the unconnected profiles on WikiTree were unlinked, but clearly a lot of WikiTreers have been working to link and even connect unlinked profiles.
https://wikitree.sdms.si/default.htm?report=srch1&Query=LastNameAtBirth%3DCrozier+needs+record&MaxProfiles=500&Format=
gives me a list of Croziers which have categories like "Needs More Records" or "Needs Birth Record", etc. applied. It does give me some false positives on some surnames when other categories happen to have the right words (or the wrong ones, depending on your perspective). It also gives me false negatives for reasons that I will have to discuss with Aleš sometime. But at least it's a lot faster than trying to go through all the possible maintenance categories looking for Croziers.