serious security problem about cyclone III device

It's not actually clear which "serious" security probelem you're addressing in your post. WP-1033 isn't particularly related to Cyclone III. It's just suggesting a method to achieve a certain degree of copy protection for designs with unencrypted configuration bitstreams. It has been shown, that unencrypted bitstreams can be decoded in principle, so it's clear, that the solution doesn't provide "strong" cryptography.

Encrypted configuration bitstreams are in constrast offering state-of-the-art cryptographic design security. Because they have a key storage inside the FPGA, they don't need devices like the said Maxim-Dallas chip.