If I'm not mistaken,
most of those empty blocks was mined very close timing to the previous block
I think it has something to do with the algorithm on picking up transactions in that short time
the miner probably has pre-set list of transactions to be included in next block,
and when two blocks are found in a very short time, the latter miner didn't have time to pick other transactions
so they removed those already included in previous block (from their current list) to avoid duplicate inclusion on current mined block
which leaves no transactions to be included if his set was the same as all included transactions in previous block
just as mentioned by ranochigo about block's validity
Given that those are likely to be miners SPV mining, inclusion of any transactions that is on the previous block would invalidate the current block, if its mined.
But aren't empty blocks mined when other miners are gathering the data of previous block.So how is it a waste of hashing power and not considered as a way to use that hash time in making profit?
It is not a waste of hashing power in term of securing bitcoin network
but it is a missed opportunity to gain extra profit from transaction fees
They prioritize the validity of mined block and block reward over extra profit from transaction fees