Skinner Engine actually "barrel bored" the cylinders of their Unaflow steam engines, achieving the same sort result the Centre Square engine cylinder had gotten by accident. Skinner's reason for this was that the Unaflow steam engine cylinder design, with the large belt of exhaust ports in the middle of the cylinder and the surrounding "exhaust belt", caused the cylinders to expand in an un-even manner. By barrel boring the cylinder so when cold, they were a bit larger in diameter at the mid section, when warmed to operating temperature the bore became a "true cylinder".
At the time James Watt was developing the rotative steam engine, the claim (if I am not mistaken) for a good fit of a piston in a cylinder was that a "worn shilling" could just fit between piston and cylinder wall. Prior to the development of the steam engine, I do not think there was any pressing need for accurate boring and turning of parts the size of engine cylinders and pistons. Early steam engine pistons did not have piston rings, but used "tow" (a kind of rope) packed in grooves in the piston rings to seal the pistons to the cylinder walls. With the very low steam pressures of that time, this worked. As pressures and temperatures increased, clearances had to tighten up and some better means of sealing the pistons had to be developed- hence better boring mills and piston rings came along.
As for the distortion of steel due to ambient conditions, I have seen this firsthand during turbine and generator erecting work. I used a very fine optical level made by Keuffel and Esser ( " K & E") known as a "Paragon" level. This worked on the principal of a surveyor's level, but had a micrometer head on the side of the instrument. Once the crosshairs were near a graduation on the "rod" (an "Invar" scale reading in 1/100ths of an inch), the micrometer dial was turned until the crosshairs cut the graduation on the rod. The micrometer dial read + or - 100 thousandths on either side of the zero, so you had to take the reading on the "rod" and add or subtract the micrometer reading. On some of the hydroelectric turbines, we'd get a large draft tube flange, perhaps 12 feet in diameter with a welded section of the draft tube attached. This had to be jigged and welded to several other plate steel sections to form a "suction bend" for the water to leave the turbine runner thru. As the welding of the additional plate sections was going on, we'd take readings with the Paragon level on the flange surface to make sure it was not being pulled too far out of flat by the welding. If it was, we'd mark the areas where that happened and the boilermakers would peen or add weld accordingly. We used to take the level readings early in the morning, since the flange and draft tube were outdoors on a temporary erecting slab. When the sun rose higher in the sky and began to shine on the flange, I could actually track the distortion by moving the micrometer dial on my Paragon level, "chasing the graduation" on the rod. When breezes blew across the flange, I also saw it change shape by using the Paragon level.
When we set the draft tubes, we used Starrett 98 levels to get things roughly levelled, and used screw jacks which were left in the concrete pour encasing the draft tubes to do the levelling. Once we had things rough levelled, we used the Paragon level for the final levelling. When the screw jacks were tweaked to final level, some additional bracing was welded to the draft tubes to keep things from shifting during the placement of the concrete. We knew it would be pointless to shoot levels on the draft tube flanges until the concrete had set for a couple of weeks and its temperature normalized to ambient.
Similarly, on steam turbine work, we used the same Paragon optical levels. We'd set the sole plates for a large steam turbine and generator on shims up on a concrete turbine pedestal foundation. The total length of the foundation might approach 75 feet, and there'd be sole plates all along the sides and front where the high pressure front standard went. We shot elevation and levels in the mornings when things were cool, and had to do it when there was little or no wind blowing as the turbine buildings were left open at that point in the job. I remember as a young engineer, being told to look thru the Paragon level and see the steel move, and being quite amazed by it.
I carry the coefficient of expansion of steel around in my head, and like to calculate approximate changes in length on things like bridges, rails, and similar "off the top of my head". In high school, in our junior year, we had a basic strength of materials course. We learned about Youngs modulus and growth due to expansion. Our teacher gave us a quiz. The problem went something like: "A vertical column made of 16" Wideflange x --- lb section is 55 feet long. If the temperature when the column was set was + 50 degrees F, and the temperature rises to + 85 degrees F, what will the compressive stress developed in the column be ?
We set off with our slide rules and pencils and paper, going like madmen. We all flunked the test handily. We had calculated the compressive stress correctly IF the top of that column were restrained. No mention was made of the top end of that column being restrained, hence, no compressive stress developed. The teacher said we'd all done our calculations correctly, but all had failed to read and understand the problem. He then told us the grade did not count, and let that be a lesson to us all. Nice guy, a great teacher and a great course which I use in my work to this very day.