The province still seems set on reducing the number of MHAs. I therefore present to the province my own proposal to help them achieve this without resulting in nearly as many problems as are possible under the current plan.
Step One: Labrador
Of the current 48 seats, they have 4. Of the proposed 38, they will still have 4. There seems general consensus that Labrador should retain 4 seats so long as the number of total seats does not drop too drastically. I'd say even with only 34 province-wide seats, Labrador having 4 seats would not be out of the question.
Step Two: Use what you got
Rather than re-drawing the entire Island-wide boundaries for Newfoundland, use what you already have. Federal ridings. There are 6 to play with. One of the simplest things you can do to help protect the process from gerrymandering, make it quicker and easier to accomplish, and allow for more focused public debate, is to begin with the 6 federal ridings.
You can then divide each of them into 5 or 6 provincial ridings, depending on weather you want a grand total of 34 or 40 seats provincially. Since 38 is closer to 40, I will proceed with that, however, 34 is fine and dandy as well.
Step Three: Decentralize.
Now that you have 6 chunks to cut into smaller chunks, you don't need a single province-wide committee to decide all the boundaries. Thus the most logical step is to create 6 committees, one for each federal riding. This means the area they are working with is far more limited, and, in effect, you can do things 6 times as fast. It's not as simplistic as that, but it will speed up the process significantly.
Step Four: Optional
This step can be skipped; but it is possible to use this opportunity to introduce electoral reform. This could work even better with 5 seats per district. You could introduce a form of multi-member STV, similar to that used in Ireland for House elections, or Australia for Senate elections. The key benefit of this is that the entire process stops here. You are done. You have all your ridings and can call the election tomorrow. These are also ridings that have already gone though the process of public consultation and approval.
Step Five: Do the work
If electoral reform is not on the table; and there is absolutely no sign it is, the next step is to delineate the ridings. With 6 independent committees doing the work, you can schedule 6 public meetings on the same day, get 6 times the feedback, and 6 times the analysis. Given how easy it is to divide something into 6 - first divide it in half, then each half into thirds - you could have basic prospective ridings ready to go within 12 hours of appointing the committees, and more thought out refined ridings within a week. This would allow plenty of time for public feedback.
Step Six: Redo if needed
Since the rush is to get all of this done in time for a spring election, there is no need to permanently keep the new ridings. They can be used for the coming election, and then the planned 2016 boundary review can go ahead as scheduled using the old process. Nothing about this proposal prevents that.
To prove how easy this is, I've done, myself, the above. You'll likely see some problems, ridings with 1.25X as many people as it's neighbour, ridings that cut some well-connected areas in half, but these are things that can be easily refined with time.