Whilst you do make a very good point, but ETCS can improve things, even slightly in that situation.
If you adopt ETCS, you have to adopt ATO if you want the full benefits from it.
ETCS will remove approach control restrictions, which avoids drivers creeping up to restrictive aspects. ETCS will allow a train to slow down a lot later. Couple that with ATO, the train can slow down much more efficiently with braking at a much later point. ATO/ETCS can eliminate some, if not a lot of defensive driving techniques.
Whilst that should be no excuse for poorly designed junctions, your still half a step forward as it will be slightly more efficient than conventional lineside signalling.
I understand what you mean. I am no signalling expert but follow the discussions in Austria, where ETCS2 (without ATO) has been in operation for a number of years on lines that are semi-HS and in mixed use (including freight). ETCS is still not seen as usable at major stations, which of course are also major capacity bottlenecks, because it would actually reduce capacity there, so the advantages tend to be limited by that. There also seems to be a consensus that even on plain line, major capacity benefits are hard to achieve. It is just that at 230 kph, cab-signalling and some amount of automation is needed and ETCS2 seems to do that job just fine.
Of course, all this could be completely different on a dedicated HSL with a uniform train fleet as well as stations used exclusively by HS services. I am on this forum mainly as an observer interested in how other countries run their railways, and as such I will be very curious how all this pans out on HS2. Especially - as stated, being no expert myself - as I gather from various sources that we should not expect capacity wonders from new signalling techniques alone. I would not mind at all if this could be disproved
Expensive though new signalling is, it will still be vastly cheaper than constantly enhancing infrastructure to accommodate more trains...