If you've already read this story, read again. Information now provided by a collector shows that the official comment was far from accurate - and may even have been created only to answer the original question, with no regard for accuracy! Additional information is in blue.
It's spot the difference time! Back in February 2014 new NCR Post and Go machines were introduced at Harpenden, and they were then rolled out over the rest of the country. One of the features of the new machines was the introduction of 'Open Value' stamps, replacing the previous 'Post and Go Labels' that were dispensed by the Wincor-Nixdorf machines.
Recognising that not all collectors had access to the machines, and seeing a way of creating another product to sell, Royal Mail made a 'Post and Go Collectors Pack' containing the basic values in Open Value format. This pack was PG15, and it was first issued on 07.07.14. Most Post and Go collectors bought one, even though the stamps were not produced from an NCR machine but from the RM2 machine C001, with datastring B7GB14-C001-1840-015.
As with the Union Flag and Spring Flowers stamps, when stock was exhausted a reprint was produced, but not announced. Can you see the difference?
It isn't obvious, and doesn't affect the pack, only the stamps:
Still not got it?
The difference isn't the same as on the Union Flag reprint, and only affects the weight -version 1 has kg and this is Kg on the reprint (version 2). The pack is no longer available from Tallents House.
A Royal Mail spokesman said:
reprints of the Union Flag, Spring Blooms and Collectors Packs were
different to the original printings because in the intervening period
a software upgrade which impacted on the appearance of the font. The
capital ‘K’ on the Machin reprint was a set up error and should not have
This is what I wrote originally:
Apart from the reader who originally drew my attention to this last year, no collectors or dealers I have spoken to are aware of the reprint. The original is most likely to be the more common, as regular customers of both Royal Mail and dealers would have got the original. Only latecomers and dealers replenishing their stocks would have had the reprint - and some dealers I spoke to still have some of their original supplies unsold! The few collectors who have only got the reprint in their collections will have no difficulty in finding the original; those who have only the original will find the reprint hard to obtain.
This is what I have since been told:
Version 2 first appeared in packs after Flags and Spring Flowers had changed; they were very rare even if you knew what to look for. I found
some at Stampex and bought some direct from Tallents House, the ratio was
about 6 times normal (version1) packs to one of the large K type (version 2).
HOWEVER. This first day cover appeared on eBay. Even allowing for post-dating - where collectors can send covers to the Special Handstamp Centres for a short period after stamps are issued (because of late delivery of the stamps or covers) - this must mean that Version 2 was available at more or less the same time as Version 1.
Even with this new information, my original comment probably remains valid. Only a small proportion was version 2 with the big K. And the early availability* would also account for the fact that the rest of the text - weight, datastring, value - is the same as on version 1.
The Big K is not so much 'Version 2' as 'Error, upper-case K'.
* The Union Flag reprint was first reported in February 2015, and the Spring Blooms in March 2015.
UPDATE 21 March
Dennis has reminded me of an earlier blog post from July 2015, when Dominic noted that in the pack he bought from Tallents House the 97p airmail stamp was printed on MA13 Machin stock. Now this makes the waters even more muddy.
If we assume that each of the values was printed on a separate roll, and then the stamps guillotined for machine-inserting into the packs (as is usual), it is fair to assume that a mix of rolls was available. The two blue rolls would be the same MA12, two of the brown rolls were undated, and the third was MA13. MA13 has not been used for either the 1L or 1LG stamps. But the 97p MA13 has only been found with a lower case 'k'. Which rather casts doubt on the suggestion that the upper case K was printed as part of the same run as the lower case. It is unlikely that Royal Mail can provide any reliable explanation of any of these differences.
Of course it makes no difference to Royal Mail at Tallents House what stamps they use, though a degree of consistency would be useful!