Explicit use of identity equations in simultaneous equation models

Question

In the Klein I model example from the systemfit package,

## Repeating the estimations of Klein's (1950) Model I
## in Greene (2003, pp. 381 and 412)
data( "KleinI" )
eqConsump <- consump ~ corpProf + corpProfLag + wages
eqInvest <- invest ~ corpProf + corpProfLag + capitalLag
eqPrivWage <- privWage ~ gnp + gnpLag + trend
inst <- ~ govExp + taxes + govWage + trend + capitalLag + corpProfLag + gnpLag
system <- list( Consumption = eqConsump, Investment = eqInvest,
PrivateWages = eqPrivWage )
# 2SLS
klein2sls <- systemfit( system, "2SLS", inst = inst, data = KleinI,
methodResidCov = "noDfCor" )
summary( klein2sls )

I would like to explicitly include identity equations (restrictions), instead of substituting them out.

The Klein I model with identity equations is here: http://davegiles.blogspot.cz/2012/05/estimating-simulating-sem.html, along with estimation of the model in EViews. I have been using this EViews example (along with the identity equations) for teaching Econometrics.

Now, I am "translating" my Econometric Labs to R and I'm a bit stuck. After searching systemfit and sem packages' help, it seems that substituting identities out is the only way to go in R.

If I missed something, please let me know.