“You have hit upon a potential mystery, which I did not explore in the book, because, by the time Megrahi went home, we had not gained disclosure of the necessary material. The Togo timer and the two control samples provided by Mebo (DP/100 & DP/111) all contained 304 crystals, yet the ones ordered from Compona were 303.”
John Ashton. Email to the author dated 08/05/2012
The disagreement between the reference number of the quartz quoted in the invoices (as well as in the 1983 SARONIX catalogue) and the quartz mounted on the MST-13 from Togo has been known to me for quite some years.
A few years ago, Swiss investigative journalist Otto Hostettler called Ulrich Lumpert and asked him about this issue while he was researching an article that he published for the BEOBATCHER in 2012. 
Lumpert told him that he really wanted the “4.194304” quartz for the MST-13 timer project because its size was convenient.
Then, last month, I decided to go straight to COMPONA AG and find out what they had to say. Again, my good friend Otto volunteered to call the owner of the company, who has been leading the company since the 80s
Otto asked about the difference between the quartz used in the Togo-timer and the one mentioned in both invoices of the quartz delivered to MEBO.
After looking at the 1985 catalogue, the owner of COMPONA, Urs Horat, confirmed that the quartz delivered to MEBO had a frequency of 4.194304 MHz. There could be no doubt about it.
Although they already relied on computer-made invoices in 1985, there was a mistake on both invoices because of an encoding error. That error was also printed in the SARONIX catalogue.
And it was just a computer glitch… Oh well, this would have been too easy!
SARONIX Quartz Data sheet (1998)