8.8.8.

Here's a few guesses at the pull to a single decimal place from measurements I've taken:

Suntour Superbe Pro Accushift 7/8 speed: 2.9 / 2.0 / 2.2 / 2.2 / 2.5 / 3 / 3.6, with a bit more (2mm?) that can be pulled past the last position and will hold as a friction shift

Suntour Accushift 3040 6 speed: 3.8 / 2.2 / 2.6 / 2.6 / 3.6, with about 5mm more of friction shifting available

Shimano Deore MT62 7 speed (hidden click at the end): 4.4 / 2.5 / 2.6 / 2.6 / 3.1 / 3.6 / 3.2

Shimano ST-EF51 7 speed: 3.6 / 2.7 / 2.5 / 2.8 / 3.0 / 3.6

Shimano Claris 2400 8 speed: 3.1 / 2.4 / 2.5 / 2.6 / 2.9 / 3.2 / 3.8

Shimano Tiagra 4500 9 speed: 2.7 / 2.3 / 2.2 / 2.4 / 2.5 / 2.7 / 2.9 / 3.7
with 105 5500 measuring pretty similar in the middle, but the first click is about 3.3 and the last is about 4.3.

Shimano R700 10 speed (same internals as 105 5600): 2.2 / 2.0 / 1.9 / 2.1 / 2.2 / 2.4 / 2.5 / 2.9 / 3.5

Shimano 5800 11 speed: a first click between 2.7 and 2.8, and then 9 clicks of 2.5mm, with less variability than all the other shifters I've measured

I did most of these with a scrap frame with a flat cable stop and a cyclecomputer magnet mounted on the cable as a reference point. Not the greatest setup but it had pretty good repeatability for the cable's absolute position. I used an 11 speed rear derailer to hold tension on the cable and the spring does seem to influence the meaurements a little so I might build a measuring tool that uses a hanging weight.

The interesting takeaway for me from some of these threads is that most shifters are very much nonlinear and it's only recently that Shimano's have become pretty linear. Only after seeing these measurements did I appreciate the clever design in SRAM's fin that changes the cable path to allow the shifter to have constant cable pull.

The other big takeaway is how many small design elements besides cog spacing and cable pull go into a shifting system with a reasonable range of working adjustment. Accushift 7/8 cable pull looks pretty similar to Shimano 9, and I have it working well on a bike (shifting 9 of 9 with that extra friction pull!), but the way the systems accomplish the shift is different, with the Accushift using more overshift before settling into position - thanks to Frank Berto's Bicycling article for the data on that.

The pull ratio for Shimano 11 seems to be more like 1.48 through the middle of the range, so if Tiagra 4700 pulls a fairly constant amount of cable like 5800 does then its pull might be a little under 2.7mm per shift to get across the 3.95mm spacing of 10 speed cogs. Looking at the positions of the shifts on Shimano 8 speed I don't think it would work. I haven't had my hands on a 4700 shifter to measure but hope to borrow one since I've never seen any data. I'd like to check out the R2000 and R3000 shifters too. If the internal design is similar to the 5800/R7000 then they should be a lot more robust than the older ones.

I have a bunch of measurements for front shifter cable pull that I took, since I've barely seen any around and it'd be useful to understand how to hack it when you need to - the way the road ones work with trim positions is all over the place, so I'll have to think about how to arrange them to best compare to the MTB. The chainring spacing is variable too.
The rough trend seems to be that the middle -> outer shift in road and MTB triples is about the same at ~8mm, but the inner -> middle is more like 11-12 for MTB and maybe 9mm for road. The older small body road shifters pulled less cable (< 1mm difference), and the new ones with long arm or toggle derailers pull a little less too. The new road triple pull is supposed to be the same as the old.