If this graph, varying from the maximum to the minimum from one day to the next, irrespective of the frequency of articles posted OR the number of subscribers interacting with the feed (I removed the numbers as the interrelation is more relevant here).
The wild swings may be explicable through “natural” forces, but the worst part is that the reported subscriber count would often drop even below the number of email subscribers. Here’s Google explaining:
FeedBurner’s subscriber count is based on an approximation of how many times your feed has been requested in a 24-hour period. Subscribers is inferred from an analysis of the many different feed readers and aggregators that retrieve this feed daily. Subscribers is not computed for browsers and bots that access your feed.
Subscribers counts are calculated by matching IP address and feed reader combinations, then using our detailed understanding of the multitude of readers, aggregators, and bots on the market to make additional inferences.
Their algorithms are secret. All I know is that the numbers change significantly, with no relation to either traffic or publishing frequency and the reported subscriber number often drops below the number of email subscribers, suggesting that the algorithm is severely flawed, especially when none of this used to happen with the same blog before Google took over Feedburner.
Sources / More info: google-help