Note where it says "In rare cases..."
It isn't that the Cloudfront access logs aren't reliable, it's just that with Cloudfront, the logs are not guaranteed to absolutely, completely, precisely, and perfectly account for each and every last request that may have been processed, nor is there a guarantee that Cloudfront will never find old, stranded log records and deliver them to you later.
S3 logging has a similar disclaimer:
The completeness and timeliness of server logging, however, is not guaranteed. The log record for a particular request might be delivered long after the request was actually processed, or it might not be delivered at all. The purpose of server logs is to give you an idea of the nature of traffic against your bucket. It is not meant to be a complete accounting of all requests. It is rare to lose log records, but server logging is not meant to be a complete accounting of all requests.
— http://docs.aws.amazon.com/AmazonS3/latest/dev/ServerLogs.html
The wording, in both cases, is not likely to be anything more significant than a simple disclaimer that AWS can refer to if a client attempts to dispute their charges based on the logs that were collected, particularly where the discrepancy between billing and logs is relatively minor.
In my experience, logs containing requests that were significantly in the past have occurred, but they are unusual, and when I have gone to look for the details of specific requests in the Cloudfront logs, they have been there.
Collect the Cloudfront logs and compare them to your detailed billing. If there is no significant inconsistency, then I would suggest that there is no need for significant concern over the implications of this disclaimer.
