Consider the ever-so-slight difference between a MIni-Roman 2D opening and a Roman 2♣ opening.
After a Min-Roman 2D opening, the usual first step for an asking sequence is 2NT. This, of course, forfeits 2NT as a natural call or as a semi-natural call. Furthermore, even if the partnership uses submarine answering, a response of 3♠ (short club) is quite a high call if the fit is hearts. Anyone experienced with using an "unknown shortness" Mini-Roman 2D opening has quickly learned how difficult these auctions are.
In contrast, in the past I used (in the context of a rather strange canape system) a 2♣ opening for a 4-4-4-1 or 5-4-4-0 pattern, with any shortness, where the range was 10-34 HCP. That range was not that difficult to use, and I remember no problems. The basic structure was for a 2D response to ask questions, Opener first showing strength in steps. Then, the cheapest call at or above 2NT asked for the stiff (sub-marine).
A simpler version was 11-15 (same range). 2D asked for strength and shape. With a minimum, Opener bid 2♥ with four hearts or 2♠ with 4144/4054/4045. With a maximum, Opener rebid one below the shortness (2NT=4441/4450, 3♣=4414/4405, 3D=4144/4054/4045, 3♥=1444/0454/0445). If Opener showed a minimum with hearts, Responder could ask for more by bidding 3D (3♥ = 4441/4450, 3♠ = 4414/4405, 3NT = 1444/0454/0445). You see that (1) Responder can bid 2NT naturally or semi-naturally, (2) the highest response is 3♥, (3) you also find out strength. Would you not rather sacrifice a 2D call (a bid that others use as the opening bid) than a 2NT call, and save that entire level of space?
Now, your system might not have this sort of capability. However, be aware of internal micro-preempts that may have negative impacts on your approach. These micro-preempts can be of much more impact than you may have recognized.