Valuation is about trying to estimate a value based on the future - and I agree this can be easier for more established companies - however when you come up against new technology firms such as those around the dot.com boom the fundamentals are still the same. It just becomes more complicated, it comes down to trying to put together a set of reasonable estimates, and sometimes this produces valuations that are way out of line with the market... and during the dot com boom this meant the market implied variables were simply unrealistic (i.e. you could instead go to the market and get the stats and find implied figures e.g. 100=.20/(X-10&#37 etc).
You are quite right about it being a multiple derived from the DDM (where price=dividend per share/(R-G)) - I like it because it breaks things down, yet remains relatively simple.
My only concern with using EV/EBITDA is that it is a firm valuation method - which is fine if you're planning an LBO but if you're only looking at it as an equity investor then you need to look at equity valuations instead... however it could be used for selection purposes by identifying potential candidates for LBO/M&A and speculating appropriately...