I had forgotten about despains declaration. Anyway I think I posted something about this maybe a week or two ago.

The 336 patent definitely does not restrict itself to a 1:1 relationship and that can only be achieved if the output of the ring oscillator is used directly as the system clock. ie if osc. clock is XMHz then system clock will be X MHz as well and not X multiplied by a certain factor and divided by a certain factor. A 1:1 is, in my opinion, very limiting and could hardly ever be used in the industry.  In the designs I have come across, a PLL or some type of a clock dividing circuitry exists between the oscillator clock and the system clock. Therefore the oscillator clock and the system clock will not change by the same amount but could change by the same ratio ie if osc. clock is 10 MHz and system clock is 50 MHz, ie a multiple of 5 of the osc.clock, and then osc. clock is changed to 20  MHz then then system clock will now become 100 MHz. Therefore the system clock changed by 50 and the osc.clock by 10 MHz and the ratio betwenn the osc. clock and the system clock is 1:5.

Let me know if that helps.