Thanks in large part to wikipedia, allcinema, tvdrama-db, and geocities, I’ve more or less finished gathering the data for my investigation of anime which first aired between 1994 and 2000 (i.e. immediately pre and post Eva). Of the 382 series myanimelist lists for that time period, 11 originally aired on non-Japanese TV, 3 were part of a pay-per-view package with no defined timeslot, 1 was a commercial series, and 1 was a bloc of shows, rather than a specific series. All those aren’t really interesting for the purpose of putting anime aired late at night in Japan in context.
I was able to find some confirmation of their initial broadcast time (either as a standalone or as part of another, longer block) for 363 of the remaining 366. 3 series, however, proved particularly troublesome to pin down. I’ve browsed through all of the above sources looking for a specific timeslot for each, and spent a healthy amount of time on google trying various combinations of keywords I’ve discovered are useful over the past week, but couldn’t dredge up that specific piece of info in their respective cases. With the likely amount of effort required to pin down that information, which effects less than 1% of the overall sample, being unduly large, I plan to simply exclude them from the final analysis. Given that, though, it’s worth mentioning which shows they are and what I *was* able to find out while looking into them.
Case #1: Burn-Up Excess
One of four anime originally to air on DirecTV Japan (the other three being the Anime Ai AwaAwaAwa block of Ebichu/Koume-chan ga Iku/Ai no Wakakusa Yama Monogatari), this one seems to have been released in chunks of 2-3 episodes on single days over a period of several months.* It may have been PPV like the other three were, but I can’t find any evidence to confirm that suspicion.**
Case #2: Ping Pong Club
I know a lot about the timeslots it reaired in: 26:50 in Kansai in 1996 and 24:40 on the Wonderful anime segment in 1997. Every source I’ve looked at agrees that it debuted in April 1995 on the TBS family of networks, but there’s no information available about its original airtime as far as I’ve seen.
Case #3: Nekojiru Gekijou Jirujiru Original
Aired as a recurring 1-minute segment on a comedy program starring the duo Bakusho Mondai – that sort of info is available in a couple different places. But said show, 爆笑問題のボスキャラ王 is a black box aside from the fact that it ran these Cat Soup episodes. Presumably, it was a program with a very short lifespan.
*Reign the Conqueror/Alexander Senki had a similar release schedule, coming out several episodes per day over four days. The difference is twofold; Reign came out over 4 consecutive days, and its timeslot is actually documented.
**I only know the other three were because the official website of the block said so.
Animetics 
Out of curiosity, what were those 11 shows that premiered outside of Japan?
You can find them on the source doc, under the “Other” tab.
I can’t see the source doc link for some reason.
Oh, it’s in the Fuji TV post. First paragraph.
Also, as a follow-up to the above, can you do an article on the sales performance of anime with relatively heavy Western involvement, in both Japan and America? For reference, a good, recent example would be Space Dandy.
There are too few series spread across too many years that fall into that category to do any meaningful analysis. It’d be like doing an analysis of college or yuri anime released in the past 5 years – you’re not likely to get any real trends, even if they exist.
Yuri might be easy to do, so maybe you can do it. Besides, somekindofthing will like it. You should probably divide yuri into two though, canon (where the yuri is confirmed) and subtext (where there’s just subtext).
My point was that there’s not enough of any of those types of shows for statistics to be particularly useful. There are what, 3 shows in the past five years that centered on a romantic relationship between two girls?
Breaking down components of a show (like subtext and side-characters) would be a hopelessly complex and subjective task because of all the other components they’d have to be weighed against.