Increase default resume block lag
Related to the recent performance issues causing observation lag on BSC - it is concerning that if large swaths of the network fall behind together (large operators), once consensus is achieved on any inbound, any restarting Bifrosts may skip over large ranges of blocks to the lag time behind last observation consensus. We should increase the allowed lag to a full day (current default is 2h) for resiliency, ensuring plenty of time for operator response to the underlying issues before any blocks may be skipped in scanning.