I'd like to raise a concern with your take on the core stealing mechanic.
In theory, Asteroid's core stealing mechanic is a very high risk/high reward way for losing teams to make a comeback. In the best case scenario, it completely reverses the teams' positions, but even if the capture fails, it creates opportunity for the losing team to reposition themselves further forward, where they can begin to gain points through killing robots.
By contrast, your core stealing mechanic is integrated into the central robot destruction mechanic, which excludes its use by teams that lack the competence and organisation required to destroy all bot groups - typically, the team that's already losing. Furthermore, destroying the last robot is no easy feat, as it's considerably beefier than the ones featured in Asteroid and it's directly in front of the defending team's spawn doors. Teams that are co-ordinated enough to pull it off are probably going to win anyway, but with how this mechanic is currently set up, teams that can destroy the final robot could easily spawn enough points from the enemy's core to win the game after completely wiping the enemy's robots a single time - all it takes is a weapon with fully automatic fire, and enough points can be spawned to win, which would of course be amplified by having multiple players attacking it.
Not only does your mechanic do away with a losing team's trump card, it seems to reward the team that manages to completely dominate the other by giving them a very high reward for the decidedly low risk act of shooting a wall.
I would suggest that for now, you either completely remove the core stealing mechanic or implement the one used in Asteroid. My long-term suggestion would be to find a way to implement the existing mechanic in a way that is separate from the main robot destruction mechanic, and that should make it more accessible to teams that are losing.