One thing to consider for calibrating a wind motor is the exact setup
of the control levers that move the slider in the tempo governor.  My
grand Pianola, at some point in its life, has had its original small
wind motor (the type with a single movable board with the two pneumatics
attached to opposite sides) replaced with the familiar see-saw design,
which is significantly larger.  It behaves just as John Phillips
describes, only being possible to calibrate accurately to the marked
tempo scale within a small Tempo range.

Thinking about why, I realised that the whole Tempo governing system is
calibrated for the original smaller motor that used less air.  Having
access to a spare governor, I took both apart and realised that the
1908 governor and the later one (from a 1920s Duo-Art) were absolutely
identical, most specifically in terms of the shaped brass plate that
modifies the air flow as the slider moves; I had speculated that they
may have used different-shaped plates to match different motors.  I
therefore measured the range of movement of the Tempo slider, and found
that in the grand it was noticeably smaller than in later instruments.
The mechanical linkage was designed to match the governor to the needs
of the motor.

In the case of something with an original wind motor such an egregious
error won't have been introduced into the system by well-meaning but
misguided fiddling.  However, you can speculate that an air motor
recovered with slightly different sizes of cloth to the original will
be affected by this in terms of its precise playing speed.  It's also
possible that the linkages may have been disturbed, or possibly never
quite right.  The point is that it's the entire system that needs
considering, not any single component.

Given all that, though, it's most likely that the regulator spring
isn't at quite the right tension, so the controlled suction supply the
Tempo slider's modifying isn't that assumed by the mechanical linkage,
which will cause over- or under-control of the speed relative to the
Tempo scale.

With regards to the Metrostyle line ...

A sharp move to the left of a Metrostyle line does indeed mean a pause,
but one of appropriately musical duration -- you have to think like
Aeolian rather than as modern slavish-follower to understand what the
line tells you!  Although these markings tend to be airily dismissed,
such an approach speaks more of modern arrogance than anything else,
otherwise why did Aeolian spend the best part of 40 years marking up
rolls in this way?  That said, the advent of the later rubber-stamped
versions does however tell you something about Aeolian's perception
(be it accuracy or staff costs) of the earlier manually-marked rolls.

You have to be a bit intelligent to use these lines, most particularly
to allow for offsets from the intended position, but there's usually a
useful signal to be obtained: try playing a Metrostyled roll and then
the same version issued under the Universal banner (an identical roll,
just unprinted) and you'll soon realise that, for those without a
firmly-set idea of how the music should go, the Metrostyle does get
you to a more musical place.

It was a clever idea as well as a worthy one: a guide for starters to
develop their own performances, which they could deviate from as they
got to know the music and had their own ideas about it.  It's a shame
that so little use is made of all the care that went into all this!

Julian Dyer