The problem is not so much the CPU speed as it is the processor's internal cache. The reason 386SX's work so well for RSS computers is because they have no internal cache to begin with.
To explain further: Older Moto DOS-based RSS (and by 'older' I mean written in the era of 286's and early 386's) was written in such a way that it required the program code to be executed in a specific sequence of steps. Later CPUs (the ones with on-board instruction caches) would, in the interests of trying to maiximize performance, execute those steps out of sequence, causing the RSS in question to crash.
This is why programs such as MoSlo have little or no effect on the problem. They do nothing to affect the CPU's cache. All they do is insert loops to slow down execution of the RSS code.
Googling for "cacheoff.com" reveals a couple of different Debug scripts you can create (under true DOS) to try and deal with this issue if you happen to be using a system with a 486-class CPU.
http://www.rigacci.org/docs/biblio/onli ... cheoff.htm
http://support.dell.com/support/edocs/d ... 000007.htm
Keep in mind that this process may differ for different systems. However, you have nothing to lose by trying. Those same pages, especially the first one, should also have instructions for reversing the process and turning the internal CPU cache back on.
Happy tweaking.
