I understand what you are saying but it does not hold up well. Everyone who added the "F" connector rework, with a better antenna (even a a crude 1/2 wave as I did), or any type of antenna improvement such as your egg beater, saw significant range increase.
IF the poor receive problems were all caused by excessively tight receiver band pass in conjunction with transmitter slot, a better antenna would not have produced the performance increase evidenced even by users who wrapped a ball of aluminumi foil around the CM15 antenna. But what you state makes my point. If X10 would have spent 20 seconds more to better tune (all) the module receivers and transmitters 10 years ago, the product would be a lot more successful than it is today. X10 shot themselves in the foot too many times by saving a penny here, a penny there, at the detriment of product performance. Greed or simply trying to stay aloat? I don't know.
I think the original design was good. Just too many "cost reductions" crapped it up.
Actually I think we are saying the same thing.
No, I still disagree (although not emphatically).
The antenna increases the signal strength seen by the receiver significantly. But the bandwidth of the receiver does not have sharp cut-off points (it's usually based on -3bB) so increasing the signal strength of a marginal signal slightly out-of-band would still give better reception.
And while it's hard to judge the skills of the antenna builders (and I disagree with the F-connector plus groundplane antenna approach), there have been those who said the antenna mod did not help but changing to the RM1SG did. Those were probably people with transmitters a little more out-of-band than those whom the antenna alone helped.
Better initial tuning of the LC controlled RF remotes probably would not have helped (and I suspect they left the factory in tune) as the frequency will shift from temperature changes and from component aging. They could have used relatively inexpensive SAW resonators (and did so for Europe) so I will agree with you that cost-cutting in the design on remotes made for N. America was at the root of the problem but not that they took a similar approach with the CM15A (which is frequency stable).
As far as CM15A cost cutting, they went outside for both the receiver and transmitter which probably cost more than reusing there oft-reused LC designs.
I suspect they went that way in order to meet European regulations which are much tighter for frequency (but allow far more power - about 10x). At the time it was designed, FCC regulations required the same testing and FCC ID whether using superhet or superregenerative so that would not have been a cost consideration. About 4-5 years ago the FCC changed its rules allowing self-testing and a
Tested to comply with FCC requirements label so there is a difference in cost now. But given X10's volume, even a few K for FCC tests might not be all that significant.
As far as I'm concerned, X10's biggest problem with the CM15A (and probably with its successor) is their absolutely atrocious software (their entire approach is idiotic) and failure to publish the communications protocol.
But, with no more RR501, no more TW523/PSC05 they have pretty well destroyed their market and I doubt a new model of the CM15A will make any difference.
A lot of people, myself included, had been using Smarthome modules/switches because they did a better job with X10 and offered a simple way to gradually shift to Insteon. I'm disappointed to see them discontinue the X10 feaures and, initially, thought it made no sense but now I suspect SH realized their products' support for the X10 protocol were helping keep X10 afloat and pulling X10 support would hasten X10's demise.
My views are colored by the fact that I had an FCC radiotelephone license in 1962 and many years later managed a multimillion dollar international operation (although we made/sold capital equipment, not inexpensive mass market consumer items).
