It's no wonder railway costs are so ludicrous then. What's wrong with a simple physical token for each section based at Bricket Wood (and if you collect it from BW, run to the terminus and back to return it, it's self-balancing) and if the sprung points are unavailable (why on earth though?), then surely it is not beyond the wit of man to have a mechanical system of some sort (i.e. similar to that used to operate the level crossing) that automatically sets and resets the points after the train passes over them? ...and why would you need a token machine at St Albans? Trains wouldn't magically spring into existence there - they would have come from Bricket Wood and thus already be in possession of the token for that section. In fact, by the same measure if the last train of the day leaves the token for Watford to BW at Watford overnight the first train can pick it up and run it to BW, and you don't need one at Watford either, eliminating the need for all the cabling.
You don't need full-on train detection, just simple common-sense rules - you can operate in the block you have the token for in either direction, and you don't proceed into the other block section unless you hold the token for it! If you are that mentally challenged that you're confused about if you physically have the token in your hand then just ring the signaller whilst you're sitting at Bricket Wood, they're nice people who should be able to get you the special help you need. The railway doesn't exist in a parallel dimension, cut off from reality, though it really seems to act like it sometimes.
Leaving aside the points for a moment.
At its simplest the line could be run with two train-staffs equipped with different annets key configurations kept in and passed through their respective key release instruments at Bricket Wood for each movement. These would need to be in a locked hut at this unmanned station. Signals or stop boards protecting single lines must have TPWS, so the stop boards here will need trainstop loops controlled by withdrawing the appropriate key, and with the appropriate indicators to confirm the loops inactive. See this example at Tenby (A NSTR remote loop).
https://www.flickr.com/photos/kevin_r_boyd/11674413745
The problem with this arrangement is what happens when one train leaves the branch by the main line connection at Watford. Unless the staff was returned by road to Bricket Wood, the second train would remain stuck at the far end of the line. The conventional way to resolve that would be to keep the staff working at the far end of the branch (hence no equipment at St Albans), but use a key token system with machines at both ends for Watford - Bricket Wood. Now the token held by the train going mainline can be put back into the machine at Watford, and another would become available for the section at Bricket Wood. That needs a cable route, as would the operational telephones at the passing point.
In simple form, a token or staff system cannot support a Penryn style loop, as one platform is in one of the single line sections. A small signalling overlay could allow it however in conjunction with staff or token, but we're still talking a little local train detection and a few signals.
Returning to the points, the original hydropneumatic train operated mechanism was a long time in development and required a lot of testing and paperwork by expensive boffins and consultants to convince safety authorities that, without a conventional facing point lock (FPL), it was sufficiently safe and secure. A FPL is normally an absolute requirement for passenger operations. The train operated mechanism was always limited to a maximum speed in either direction of 15MPH nonetheless. All the train operated points in South Wales, including Tenby in the photo linked above and the Central Wales line, have now been replaced by motor worked points (I checked the sectional appendix: there's a note to the effect for every loop). Whether train operated or motor operated a set of points needs an indicator in the facing direction, and with motor operation an indicator in the trailing direction too, which can be combined with the stop board TPWS indicator. All these indicators are starting to look like so many colour light signals, and with the motor operated points requiring some local train detection to make sure they're clear of trains before moving it's all starting to look a lot like a fully signalled layout.
There are other simple point mechanisms available from European manufacturers used for light rail and tramways, but as far as I know nothing currently approved for UK heavy rail passenger use. If a project such as Watford - St. Albans wished to use such a new product they would have to take on the cost and risk of getting that approval (with all the expensive boffins and consultants involved).
TCB using axle counters only requires one rail mounted inductive sensor head unit at each end of the single line regardless of the length of the section between, so full train detection is no longer so equipment and power intensive as it was with track circuits which are length limited to typically around 1km. Yes communications cables are required between the two ends of a counter section, requiring a cable route, but you'd need a cable route for a token circuit and telephones too.
I think token or staff working to modern standards would offer no real savings over a modern lightweight modular signalling installation, yet would slow down the passing loop operation and require drivers to leave the cab whilst exchanging keys. Even a token and staff system would have to be supervised and controlled by a signaller somewhere, so if added to the WCML workstations it would still require a change to the screens and controls.
Here's a useful page on the subject:
http://www.railsigns.uk/photos/p_points1/p_points1.html
