It's a matter of physics for me.
A fishing boat is a simple steel hull which doesn't have much of an acoustic signature apart from the diesel engine.
It is much more than physics. Each vessel wil have a unique set of acoustic anomalies and thus have a unique signature. In order to get all of that, you would have to have a very sophisticated program to mimic them.
So they could strip the fishing vessel, switch the engine off or dampen it, and then add a bunch of other frequencies with varying volumes and durations.
Then it's a question of how close a submarine has to get it has to get, before it can determine if it is a fishing boat or Type-56.
Sheer numbers will complicate the situation immensely, and we're talking only talking say a total of $400 million for 1000 fishing boats. That is dirt cheap..
Sorry but the shear scale of what you are proposing is monumental.
Basically either dismantling and refitting hundreds of small fishing boats, or purpose building new decoys to those very specific acoustic signatures.
Then you have to deploy them and control them appropriately, while at the same time allowing for all the "real" fishing boats and other commercial traffic.
Even then, unless you are thinking of deploying these over the entirety of the China Sea, an SSN or a Japanese or Vietnamese SSK will do one of three things:
1) Identify your decoys for what they are because they do not match any specific or even general signature.
2) If they cannot ID them, then simply maneuver around them and come at the target area from a different direction.
3) If the specific target of the mission is unable to be reached, then "Charlie Mike," to the alternate area of operation entirely.
Again, for the PLAN to successfully do what you are proposing over large segments of the China Sea would be much more expensive, and much more technically challenging than you are suggesting here. And it would be a logistical nightmare to the point of not being able to be effectively done.
I believe there are some decoys that could be employed in very specific circumstances to protect point targets of interest...but even then, it is a sophisticated effort to try and mimic what the US or other nation already has specifically recorded as a specific Type 056 signature...because that is what we are talking about here, the individual vessel's signature because they all have their unique anomalies...and that is what the US is constantly doing, tagging the unique characteristics of each vessel (Corvette, FFG, DDG, SSN, SSK, LPD, etc.) they come across and then building their data base accordingly.
1st it has to match the very specific information they have for the Type 056 in general, and the US and Japanese are very, very good, and have decades of experience doing this. If the "decoy," gets past that (and that is not an easy task in the least), then they will see specifically which vessel it is based on the data base they have. They know the build rates and what is being constructed just as we do...in fact, much better. So, very quickly they will catch on if it matches nothing they have, and if there are so many of them, unless they can make it mimic the actual acoustic anomalies of a specific vessel...which is, as I have indicated, very, very difficult to do.
So, all of that would have to be taken into account. A very difficult thing to do over they types of area you are proposing.
IMHO, a better approach would be to try and have a much smaller number of specific decoys for specific "targets" the PLAN wants to protect and address them on a more specific basis.