Saga of the perlexing swarm

About a month ago, April 9, a swarm flew through the yard, clustering high in a tree behind us.

The cluster is in the center, albeit difficult to see

I put out a bait box, but the bees stayed put, perhaps due to the inclement weather.  After a couple of weeks had passed, I concluded that the swarm was not going anywhere and would build an exposed hive, outside of a cavity.
Then, on April 29, the bees left the tree, flew around for about a half-hour and the clustered on a pole of my grape vines.

After landing on the pole

A closer view

I saw several bees looking at the bait box, so my hopes were renewed that they would move in by their on accord.  I waited a week.  The bees formed a more compact cluster and, subjectively, it looked smaller than when it was on the tree.

The day after clustering

A week later

I was of two minds: observe and let nature take its course, or intervene and try to save the swarm.  Yesterday I elected to take the latter course.  I put a nuc box beneath the cluster and shook and brushed the bees into the box.  There was a 5-inch comb with nectar in it attached to the pole which I broke off and put into the nuc. Initially, I thought I had missed the queen because many of the bees went back onto the pole.  I put frames into the nuc and left it uncovered by the pole.  A couple of hours later, most of the bees were in the box, a small bunch on the pole.

A few bees stayed on the pole

The bees seem to have settled into the nuc.  I put the nuc on one of Papa Eddie's tables that needed only a little re-enforcing.

The nuc beneath the bait box

A worker exploring the entrance of her new home

I inspected hive #1, 2 & 3 today.  There is a lot of capped worker brood.  The bees only cover 5-6 frames, so I did not add a second box today.  The brood pattern looked overall good, but there was a suggestion of some spottiness.  I put the sticky boards in today, however, I think I will switch to alcohol washes as a means of monitoring mites, something that Randy Oliver suggests is more accurate.