I've decided to bite the bullet and convert from Imperial to Metric.  One item I'm wondering if it's really necessary to convert is the Incra Master Lift II.  Is there ever a time when it would be critical to have the Metric setup versus Imperial? 

In the Imperial setup, there are 32 increments per revolution, each increment being .002 inches and 1 revolution being 1/16 inches.  In the Metric setup, there are 40 increments per revolution, each increment being .05mm and 1 revolution being 2mm. 

Since .05mm is roughly .002 inches, is it really necessary to upgrade the lift to Metric for $60?

They advised me to call JessEm who makes the lift. JessEm said it was $30-40 plus $15 shipping so it will be $45-55 for new lead screw and dial.

I'm with Peter, a digital height gauge is the best way to go,also use mine for setting offset to the bit and fence by lying the unit on it's side.  I use it for my table saw height and offset as well.  Another useful too is digital calipers, I have 4,6,and 8" units.  I just used my 8" unit yesterday to set my new TSC 55 to match my corded 55 and 75's 3mm offset on my rails.

I just used the digital gauge and it does appear every increment is .002 as advertised. While the increments don’t correspond directly to metric, there isn’t a positive locking mechanism between increments so I can lock the height between increments allowing for increments of .05mm. I could potentially achieve even less but my digital scale doesn’t measure any more precise than that.

Actually if you do 10 increments on Metric but only 9 on Imperial, you have a difference of .00169.  And that is full increments yet you’re not required to do full increments so you could technically get them very close to being the same; however, I’ve not seen a digital height gauge which measures less than .002” so perhaps it’s not that critical in woodworking?

There are a significant number of variables that will impact accuracy beyond the increment count and/or splitting increments. The proof is in the cut and checking the setup in total. Trial cuts are what I rely on. I find nothing wrong with trying to ring out as much accuracy and as high a resolution in the mechanisms or digital scales. The more you control for variation the better off you are. Any setup on any machine is subject to dynamic variables that can change results from what we expect to get.

Even though I disagreed with Pat Warner regarding the use of router lifts (he eschewed them) I respected the analytical degree to which subjected his equipment. Dial indicators resided on each side of the fence, all components where optimized to remove the unknowns. Good setup hygiene. I believe he disregarded lifts because he felt they introduced too many opportunities for setup variation and deterioration. That may have been the case a decade ago but lifts and their accessories have been continuously improving and their ease of use and speed of use have, IMO, overtaken the limitations and weaknesses he saw in them.