As a developer, I have to respect this.  You guys have been honest since day-one and have stated the fact that Lisk is a fork and improvement upon Crypti.  As such, I imagine you guys have a long laundry list of changes that you want to make.  Some can and should be made before launch while others can be put on the back burner for post-launch updates.

I also can understand why the team may have initially ignored updating the DB infrastructure.  It's not a small change.  If the original Crypti dev decided to use SQLLite, it has proven that SQLLite can work on a project with limited success.

This type of change/update proves that the Lisk devs want the launch to go as smoothly as possible.  With limited testnet usage (only the hardcore supporters are running test nodes ATM) then it is important to make sure that once the official release takes place, the underlying DB can handle the mass influx of new transactions.

Apparently after launch, and during it's peak usage, Crypti didn't experience any problems with the SQLLite db.  That is good.  However, if LiskHQ (and the other devs) had a bad "feeling" about sticking with the current tech, I 100% support the delays so that they can update the back-end db TODAY, rather than later when usage spikes and the blockchain starts crumbling from excessive load.

TL;DR

LiskHQ and the other devs are making a wise decision to fix a problem before it becomes a legitimate issue.  They are exercising foresight in taking care of Lisk, rather than waiting for bugs to become issues that bring the blockchain to a halt until a *cough* rushed *cough* patch has to be released.