Hello,
I am Milos and I am bureaucrat on sr:
First of all, I have to tell you some introduction in our alphabet problems.
Until the second half of 19th century Serbs used only Cyrillic alphabet. (In the ancient history, Serbs was using Latin and Glagolic alphabet for a short time, but without stronger cultural consequences.) Between the second half of 19th century and the end of The Second World War, Latin alphabet was not used a lot; it was used by very small number of writers. After the end of WWII, strong cultural influence of Communist Party of Yugoslavia gave to Serbs active usage of Latin alphabet. During 1980's and 1990's Latin alphabet was used more often then Cyrillic (maybe 60%, maybe even 70%). Today, the situation is something about 50-50 with tendency of Cyrillic usage growth. In this moment we have very strong xenophobic movement. A lot of them are very active on Internet and on Wikipedia, too. They do not accept Latin alphabet as "Serbian". Even a lot of people use Latin alphabet.
I made voting for the question: Would Serbian Wikipedia has Latin pages too? Even the end of August is the end of voting for that question, I think that usage of Latin alphabet would not pass. Situation is 4:2 against and Serbian Wikipedia doesn't have a lot of active users (we have one or two new users which can't vote).
Even I use Cyrillic alphabet, I relies that usage of Latin alphabet is important for us: We have significant minority (about 50%!) of Latin users and if anyone starts to search on Internet, (s)he uses Latin alphabet, not Cyrillic (usually even without Serbian Latin letters). We have very ironic situation: If you type the name of one of Serbian 19th century writers, Radoje Domanovic (in Serbian Cyrillic: Радоје Домановић, in Serbian Latin: Radoje Domanović) in Google, you will get: (1) Cyrillic: Serbian Wikipedia is on the first place; (2) Any Latin: Wikipedia doesn't exists.
So, I think that we can do something like this: We can make secondary Serbian Wikipedia using secondary two-letters code for Serbian language: sp. It would not be "real" Wikipedia, because it should be used only for transliteration of Cyrillic Wikipedia (transliteration from Serbian Cyrillic to Serbian Latin is algorithmic, the reverse process is fuzzy). All of the pages on sp: should be "protected" and when someone writes something on Cyrillic Wikipedia, it should be transliterated to Latin, too. Also, all of users should be the same.
I think it is good solution because we would not have multiplication of work (such as we have now: Croatian, Bosnian and Serbo-Croatian Wikipedias are using the (very, very) almost the same language). All of people would work on sr:, not on sp:.
Also, I have some notes to developers (if the request passes): about technical solution for that. It would not any consequence to Wikimedia's performances because it is (almost) row transliteration.
Best, Milos