If so, how come the first 4-5 CuePoints get successfully injected into my flv?
There are no other cue points/ markers at the same position, so there is no logic conflict. Otherwise of course there are limitations on handling such data, both in AE and Flash like AE only embedding the markers of the topmost layer in teh comp (if there are no comp markers, otehrwise those take precedence). Your workflow simply doesn't make sense.
Is there any other way?
Parse the data from an text/XML file, which I was suggesting from the start.
"There are no other cue points/ markers at the same position, so there is no logic conflict."
Actually there is. I checked my flv file and it contains 4 CuePoints. The first one is the only CP at its position but number 2 and 3 (out of my 4 CPs) are at the same position. Thats why i was wondering.
Anyway, I found another way which hopefully works. I render the flv file once for every layer that has corner pin data. So i open my project, remove all layers except my footage and one (out of 5) layer that contains corner pin data and render it. I open the rendered flv file in RichFLV and export the CuePoints from that layer as xml. Once this is done for all the layers, i put together all the xml files to get one xml file that contains all CuePoints and then i open my flv file once again in RichFLV, delete its CuePoints (cause they are not complete but just the Cuepoints of one layer) and then import my xml file that contains all CuePoints of every layer and apply it to the flv file .
@Mylenium: Thanks for your help