I thought I'd research this online a bit this morning.
The 'ban' on passengers (and catering crew) from the whole of the leading vehicle, as written in the railway group standards of the day was negotiable with HMRI. This is often the case when the actual text is found.
(Compare with the regular posts where 'banned' is used rather than 'not yet cleared' to justify the inability to use certain stock wherever.)
Interestingly, the oft quoted '⅓ rule' cannot be found in the standards either; they went directly from no passengers or catering staff being allowed in the leading vehicle, to the complete paragraph being removed from the standard. The dates of the various issues quoted definitely imply that it was the modern fast 125 mph EMU/DMU classes that caused the removal of the requirement. The text as written would have meant an empty carriage at both ends of a four car 125 mph DMU of class 22X? So the minimum train length would have had to be at least 2+4 or 5 for useful purposes. IMHO the important thing is that a single unmanned DVT was probably an acceptable idea on the other end to a loco, but the move towards distributed traction makes a DVT on both ends start to look very uneconomical and indeed a bit ridiculous.
So going back to RGS GM/RT 2100 Issue 1 of 1994, and Issue 2 of 1997
you find this paragraph:
4.8
Passengers and catering staff shall not be carried in the leading vehicle of a train which has a maximum operating speed greater than 160 km/h
without the sanction of HM Railway Inspectorate. A quantified risk assessment shall be submitted in support of any proposal to deviate from this requirement.
(My bolding)
http://www.rgsonline.co.uk/Railway_...ck/Railway Group Standards/GMRT2100 Iss 2.pdf
The next issue of the RGS, version 3 of Oct 2000,
completely removes paragraph 4.8, and by Issue 4 section 3 a re-arrangement of sections brings in paragraph 3:
3.3.1.1 The structural crashworthiness requirements of BS EN 15227:2008 shall apply. The collision scenarios set out in section 5 of BS EN 15227:2008 shall be applied in accordance with the crashworthiness design categories set out in section 4 of BS EN 15227:2008.
http://www.rgsonline.co.uk/Railway_...ck/Railway Group Standards/GMRT2100 Iss 4.pdf
Issue 5 (current) 2012 is fairly similar to issue 4. With no access to British Standards we cannot see the text, but I believe the EN means they are a European harmonised standard; and we know that passengers in leading vehicles are common in modern European stock.
I suggest therefore that the amendment dates of the RGSs show that the designers/builders of the 22X and 395 must have gone to HMRI and negotiated for, and had accepted, a proposal that passengers would not travel in the leading ⅓ of the driving vehicle, that seats would be mostly back to direction of travel in the leading car, and significantly they must also have gained an acceptance that catering staff could now travel nearer the driver in safety.
HMRI having accepted that design for new classes, the RSSB then took the pragmatic view to remove the restriction on anyone travelling anywhere in the leading carriage.
Please note, I haven't linked to all 5 versions of the RGS that I mention - if anyone wants to read them just change the final digit in the URLs as necessary, i.e. 1-5.