The output of spot finding shows the number of strong pixels found on every image and then at the end shows an overall summary:
Extracted 405471 spots
Removed 123788 spots with size < 3 pixels
Removed 34 spots with size > 1000 pixels
Calculated 281649 spot centroids
Calculated 281649 spot intensities
Filtered 234130 of 281649 spots by peak-centroid distance
Histogram of per-image spot count for imageset 0:
234130 spots found on 1800 images (max 4530 / bin)
* ******
******************** ********** ***
*************************** ************************ *******
************************************************************
************************************************************
************************************************************
************************************************************
************************************************************
************************************************************
************************************************************
1 image 1800
--------------------------------------------------------------------------------
Saved 234130 reflections to strong.refl
This "ascii-art" graph can be useful to give you an overall indication of how the data look. If, as here, the graph is approximately constant you may assume that the sample has not suffered substantial radiation damage and was reasonably well centred. If there are sinusoidal variations this may indicate either a variation in diffraction strength (e.g. anisotropy) or a significant variation in the unit cell lengths (i.e. when a long axis is perpendicular to the beam there are more spots.)
This summary also shows how many reflections were excluded for being
too small or too large - usually the defaults are sensible however,
but you can adjust these with min_spot_size and max_spot_size.
This is also interesting:
Filtered 234130 of 281649 spots by peak-centroid distance
If you look at the images you will see in the default spot finding that there are quite a few "missed" reflections - spots wihich were not found. By default the spot finding in DIALS also filters on the distance between the highest pixel and the centre of mass (with the max_separation parameter). In this case a lot of spots have failed this test, as in fact the crystal was split: this is visible in the reciprical lattice viewer, by looking at the unindexed reflections which appear to fall along lines of indexed reflections.
If we relax the max_separation parameter then many more spots are found but the number of indexed reflections will not increase that much: allowing a second lattice to be indexed increases the number of indexed reflections but - importantly - not usefully. The crystal lattice is still essentially "single" with more complex reflection profiles which are actually handled correctly in integration (i.e. all of the intensity is measured) hence it is probably better at this stage to ignore the details and continue.
Actually looking at the images:
The shoeboxes are sufficiently large that all the intensity will be measured even if the spot is split, though it could be the case that the data are better integrated by simple summation than profile fitting.