This gentleman has brought forth some excellent videos; these bring a general awareness of what processes for one's engine in the way of machine work that might be advised, including an introduction to the actual process involved, that many might not be ware of. I say bravo for this effort!

Now I do, and so should others, understand that "there's more than one way to skin a cat", and that different equipment responds differently and therefore the specified execution may be different in one instance from another in order for the same out come, and I always say: "If it's working for ya' keep at it", but . . . . . . In this "Line Hone Video" I'm going to comment in an attempt to provide some supplemental input for those whom might be interested.

First thought, I've noticed this here and in other videos of his, when measuring other stuff, but it seems he needs to buy a dial-bore gauge that actually reads in 10th's, not half thousandths. This so that the instrument is actually "telling" him where the dimension "is" in 10th's, rather than him "reading into" the measurements and even what one might want to see. It also gives one a better perspective of what's happening as going along in the process, this before too great of a correction is needed due to the numbness in the coarser read-out. And here, 10th's do matter.
Next, now each block 'is' different, and hence the response in the relationship to the machining efforts can and will vary, but I've line-honed a lot of blocks, including a lot of "FE's", this including the training of apprenticed individuals, and in watching the video and the results, which might seem consistent with what are poor practices and perhaps questionable results.
For example: there seemed to be an excessive number of stokes commensurate with the sum of material removed; either the stones are of the wrong material makeup for the block, or they are "loaded-up" and "fouled", and/or the "pressure is set to low". The fact that the block side, being the lower side/hemisphere "cleaned-up" readily, particularly so early in the process leaving the caps to such a degree of "unfinished" at the same time isn't as intended or as is generally consistent with this process when properly executed. Though this is as of expected if the pressure is set too low and the weight of the mandrel becomes overly influential in which surfaces are being cut. The unappreciated result is that greater sums of material are being removed from the block saddles vs. the caps in the endeavor to bring the bores back to size; this being as presented in the video; and by the need to "go at it again" (and again) in order to chase the "shadowing" down in the caps to an acceptable degree, but unfortunately results in moving the main bore centers up in the block excessively than perhaps otherwise would have been needed to be experienced.
As in reality, it is most often the lesser mass of the caps, which are actually enduring the greater load (as the crankshaft is being pushed downward against the caps by the load created of the firing cylinders above) and are what is generally getting "out of shape" to the greater degree. So consider that in the preparatory work before the honing mandrel approaches the block, the caps are cut down a sum at the parting line. This reduces the diameter of this hemisphere to something less than that of the unaddressed hemisphere of the block, and this smaller radius is the surface intended to be cut as the honing process attempts to make the hole round again. But this cap bearing saddle radius now is of an eccentric manor and this brings up the reason for the often unaddressable surfaces at the parting line.
Now in some instances, as voiced in the video, the blocks may "get out of alignment" due to use and just heat-cycling and this usually leads to the "banana" or a "bowing" progression result in the main bore alignments vs singular bulkhead displacements, which are more often the result of other item failures (loose rod, broken crank, etc.). So in the initial honing process the individual bore roundness geometry may actually get worse, as the mandrel is possibly removing material from only specific points in the circumferences, averaging the relationships in current bore positioning, on the way to bringing them back to the desired size. And it is important for the operator to be able to "read" this, so that they will have a good feel for what is going to be required to achieve success and allowing for a minimal sum of material being removed. But in the video, for example in the case of the #4 cap ("the worst one") not cleaning-up squarely (as seen @12:22 time point), the statement that it "was leaning" is correct, but based on the evidential pattern left from the hone, this result is generally caused by the inaccurate cutting process of the cap, as not having been parallel to the existing surface as from the O.E.M., aka. a self inflicted wound in the "fix-it" process! And this can be particularly problematic with the thrust main (#3 here) as it will cause the thrust bearing faces to be tilted, resulting in a loss of bearing face squareness to crankshaft surface and a loss of "end-thrust" value.
Also, one should "never" enlarge the bores beyond that which you wish to finish them at. Though this is often practiced to cleanse the saddle bores at the parting lines, this to better appease the customer in his critique of the job, then requiring a follow-up redo to bring the bores back down to size, but although one might accomplish getting the vertical dimension "in" size, the "X" dimensions (2 - 8 & 10-4) and anything closer to the parting lines just got worse! But, if you wanted "even more" eccentricity in the bearing oil clearance factored in . . . . .
This "roundness" vs. just the "vertical" dimension is also a concern when one hones continuously from one end of the block, and an example of a negative effect is also presented in the video when the observed measurements stepped from larger to smaller end to end in the block. Now this was identified in the video, but better results are had if one doesn't wait until the differentials are so great (here a better dial bore gauge might help); it's better practice to rotate the block more frequently in an attempt to keep the bores close to one another throughout the process as each bore does present an influential effect on the others, as it is a constant case of averaging together the sum of all. Also "flipping the block" is necessary as the makeup of the mandrel and how it functions in the removal of material process, this in what as prepared are more or less oval holes, as it tends to "drive" off-center as it works; so again, one can get the "vertical" measurement "in", but if checking the "X" dimensions, you might be somewhat disappointed! And then, though somewhat unimportant to the customer, is that flipping the block aids in keeping the mandrel from developing a "pattern" or uneven wear in its' length, which it then attempts to transfer to the next job, this resulting in making it more difficult for the machinist to control the outcome in future endeavors.
Even if none of the above had any relevance as having been suggested, and no matter what as presented, nothing changed the outcome from that as shown in the video, then in order to influence the mandrel to cut the caps versus the block saddles as seemed to be taking place, something else, though unconventional, but what has worked as practiced by myself and knowing others, is to flip the block over, main caps down! Adding the weight of the mandrel in the mix will increase bias and aid the stones in biting the caps better. As remember, it is the "undersized" caps in which you really want to remove the material from, not the already to (or beyond) size block saddles, not mention creating a sloppy timing chain result. And as relevant examples, this is what is often practiced when line-honing O.E.M. iron blocks with "billet steel" caps installed and aluminum blocks, particularly those with the dissimilar ferrous metal caps, as in both instances those caps obviously will resist the cutting action as compared with the softer material of the block.
Just "food for thought"; though there 'is' more on this topic for consideration.

Scott.