not really..
what you will see is that when there is an "empty" block. its not due to naferious intensions of leaving a block empty..
instead its due to the fact that normally it takes upto 2 minutes to validate a previous block and then start hashing a new block of unconfirmed data.
some pools know that the risk of orphan is just 2%. so wasting 20% of a 10minute slot validating data for a 2% risk. is something that they think is not financial savvi..
the issue is this though..
they get lucky.. they solve a block in under 2 minutes before they had a chance to add new unconfirmed transactions to that block try.
its not intentional ignoring transactions. its instead not hanging around hashing nothing for 2 minutes..
as for the GFW.. thats not a big deal.. the pools(holding the blockchain data) are outside of the firewall so no issues talking to nodes. and its only the ASICS (that dont see the full blockchain) that are inside the firewall..
It may not be for nefarious intentions but with such a majority of hashing power between these pools it's something that actually does occur and is completely unfair to miners that are not peering with them directly. Even if you were peering with them directly you would only be relying on trust, that your peer is not withholding a block from you.
Also how do we know the pools(assuming you're referring to the bitcoind instances themselves?) are for certain outside of mainland China?
What about the following stratum servers? You think they may be connecting to bitcoind instances outside of the country?
f2pool stratum.f2pool.com 42.96.248.230
AS37963 Hangzhou Alibaba Advertising Co.,Ltd.
Country of Orgin: China
BTCC stratum.btcchina.com 180.97.161.9
AS4134 China Telecom Backbone
Country of Orgin: China
BW.com stratum.bw.com 120.55.153.228
AS37963 Hangzhou Alibaba Advertising Co.,Ltd.
Country of Orgin: China