I'm mostly speculating, but it boils down to profit and meeting the demands of the evolving market and competition. 

As for profit, the basis of the Unimatic mechanism is this huge piece of cast iron - the mechanism base.  It probably weighs in at 10 pounds.  The casting and machining costs alone on this would be enough to force the change in a high volume/domestic appliance, let alone the cast iron found in the rotating mass that produces the pulsation for agitation.   And while it could be argued that a two speed motor could have been incorporated within the mechanism the fact that the rotating speed was 1140, and not 1725, would have forced a change within the gearing for agitation if multiple speeds were desired.  This is also before considering the amount of torque required at the start of spin.  It would have been costly to make this work.

As unpopular as this opinion may be, the Multimatic mechanism was the right move for the time, even with the many issues that came along with the '59 models.  Most of the bugs were worked out by '60 and by '63 the overall clutch drive and transmission were improved enough that it was put into commercial use by '65. 

The Unimatic serving the commercial side of the business in 1959 was also a smart move.  New tooling wouldn't be needed to support the platform, multi-speed washers were not a requirement within a commercial setting, and the overall mechanism was much more robust to take the abuse of hour over hour use throughout a day.

While the Rollermatic mechanism may seem and feel more cheap compared to either the Unimatic or Multimatic, it also shows how over built both platforms were by the time GM introduced the Rollermatic, that had been in development for almost 5 years by the time it was debuted in 1964. 

My best guess is GM would have ditched the 1-18 Rollermatic mechanism by the mid-80s and had gone with an agi-tub / agitator based washer platform, before selling off the Frigidaire division to the highest bidder by the end of the 80's.

Ben