I saw a similar procedure described on a Subaru site.
I think it's called "torque to yield".
Anyway, the old tried and true method certainly seems easier to use than the methods specified for some newer engines.

The new method has to do with eliminating the variables of friction losses and probably some other variables, and instead is calibrated using the known strengths of all the materials involved. Apparently the bolts get temporarily stretched when tightened and so are under tension - until they are pulled past their elasticity, then they are permanently stretched and the resulting pressure on the mated parts is now different.

I think it also has to do with the newer aluminum blocks with steel or iron cylinder sleeves. Subaru and apparently most other makers had plenty problems with head gasket failure in these so there have been changes in head gasket materials and the tightening of head bolts.

--
Sven: '89 245 NA, 951 ECU, open-front airbox, E-fan, 205/65-15's, IPD sways, E-Codes, amber front corner reflectors, quad horns. Wifemobile '89 245 NA stock. 90 244 NA spare, runs.