While I do like the idea of shooting things instead of shining pretty lights at it (that's why I prefer hybrid weapons over lasers), something bothers me about this system as presented in the blog. Probably somebody has pointed it out already, but reiteration is always good on the Interwebs, isn't it?

So from what I gather the structure enters the vulnerable timer at a predetermined time but the repair timer starts the moment it gets shot at and will repair the structure to full after the timer runs out. Even if it has been damaged in the previous vulnerability phase.

There are two problems I see with it.

First, there is no real reason to have this vulnerable state going on longer than the repair cycle since all the defender has to do is shoot his own structure once at the beginning (with an out of alliance char if necessary) to start the repair timer and the structure will be as good as new after that relatively short time. An attacker planning to attack near the middle or end of the vulnerable timer will have to start anew. To be fair the current entosis mechanic has a similar consequence that not being on field the moment the vulnerability starts puts you at a potentially huge disadvantage but it still takes time and effort to capture nodes distributed over several systems. It's far more difficult to block an attacker from several systems than from just the system the station is in.
Shooting your own structure to help repair it faster seems counter-intuitive to me...
There are two ways to address this: If you want to have both attacker and defender to be present on the field the moment the new vulnerability starts just start the damn thing in the repair phase right away. If you do want a bit of flexibility for the attacker to properly deploy at any time or break through a gatecamp during the vulnerability window apply the repair to previously shot down HP pools at the end of the vulnerability window regardless if a repair timer has run it's course or not and allow multiple repair cycles to occur during a single vulnerability period that only deal with damage to the current pool.

Second, I don't think it's right that everything is repaired. What a repair cycle should do is either repair the damage done to the current vulnerable pool or the one before it - and only that one. So let's say the structure is in structure vulnerable mode, survives a vulnerability period uncontested it doesn't get it's shield and armour repped to full but just the armour. It will need a second vulnerability window to get repped to full. To be fair to prevent delaying tactics by a single player (or small force) that has no real hope to actually apply enough dps to be a threat I would like to see some sort of threshold. For example, as the damage to the current pool doesn't exceed - say - 25% the next pool gets repped too after the repair cycle. Or you could just go with repping the current and previous pool although than this is only really relevant if the structure is already in structure.