My view is that the vast majority of SAAS companies should...

My view is that the vast majority of SAAS companies should forcibly expense all R&D (as OCL does) because the usable life on newly created software is very short - almost certainly less than 5 years and quite often only a year or two. OCL itself releases a major update on its ECM package every year or two (which is what drives the large R&D expense), so there's no sensible argument for capitalising that expense. The fact other software companies choose to take advantage of flexibility in accounting rules to capitalise software R&D isn't really relevant, i think - at the end of the day it's all cash out the door howsoever it's defined.

My ~23x EBITDA already does make the adjustment for OCL's cash balance, i.e.: ($267m market cap - $21m cash) = $246m EV / $10.5m EBITDA = 23.4x. If i took pre tax operating cash flow rather than EBITDA, then the ratio is $246/$11.3 = 21.8x, or a c.4.5% pre-tax cash yield to EV. If i cash-adjusted the PE for the $0.20 per share net cash balance as you suggest, then the PE would be ($2.89 - $0.20) = $2.69/$0.08 = ~34x, which is still very elevated.

Whichever of those you choose to take, at a ~4.5% pre-tax or ~3.2% post tax yield, the stock is still very fully valued and inherently assumes there's a lot of growth into FY19-20. If that growth doesn't come through, then there's no good reason why the stock should continue trading at >20x EBITDA and >30x P/Es - it should de-rate to something more like 12-14x EBITDA and 20x P/E. Something clearly changed quite significantly in 2H18 as the outlook given at 1H18 was very positive, so to my mind it's a little brave to underwrite significant growth into FY19-20 just as OCL's customers (state and local governments, primarily) are about to face significant budgetary constraints due to the residential property market downturn.