Why does it make sense for mining pools to send empty blocks to miners in Stratum v1?

Question

I am reading the recently released specs of Stratum v1 versus Stratum v2. One of the features is that "it is equally as efficient for pools to send full blocks for miners to begin working on as it is for them to send empty blocks ". I don't understand why it is efficient for mining pools to send empty blocks with Stratum v1 and at what stage of the mining process they send empty blocks. So this boils down to two questions:

1. Why is it efficient for mining pools to send empty blocks with the current Stratum protocol?
2. At what stage of the mining process do they send empty blocks? Is it when they are waiting for a miner to submit a block with valid proof-of-work?

Answer 1

Historically, block template generation has been very slow. It was faster to send an empty template with no transactions, then wait for the mining pool to create a block template with transactions, then send the result to the miners when it was ready (in the order of ten seconds occasionally). This is no longer the case in modern versions of bitcoin core.